Device management method and system based on Internet of Things

Through the Internet of Things technology, the charging current value of the charging pile is collected and analyzed by current sensors, real-time monitoring and fault diagnosis of the charging pile is achieved, the problem of inefficient traditional manual inspection is solved, and the reliability and user of the charging pile are improved. Experience.

CN118444051BActive Publication Date: 2024-12-03SHENZHEN GUOHUI METROLOGY QUALITY INSPECTION CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202410561059.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-08
Publication Date
2024-12-03
Estimated Expiration
2044-05-08

AI Technical Summary

Technical Problem

Traditional charging pile management solutions rely on manual inspection and are inefficient and cannot monitor the state of charging piles in real time, resulting in delays in troubleshooting and affecting user experience.

Method used

The charging current value of the charging pile is collected through the current sensor, and timing characteristics are analyzed. The Internet of Things system is used to achieve remote management and fault diagnosis, improving the reliability and user experience of the charging pile.

Benefits of technology

Real-time monitoring and fault prediction and maintenance of charging piles are realized, avoiding the problems of low efficiency and low accuracy of manual inspections, and improving the reliability and user experience of charging piles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118444051B_ABST
    Figure CN118444051B_ABST
Patent Text Reader

Abstract

Disclosed is a device management method and system based on the Internet of Things. First, the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor are obtained, and then the charging current values ​​at the multiple predetermined time points are arranged according to the time dimension as a charging current timing input vector, and then the current timing characteristic analysis is performed on the charging current timing input vector to obtain the charging current timing characteristic, and finally, based on the charging current timing characteristic, it is determined whether the working state of the monitored charging pile is abnormal. In this way, the reliability of the charging pile and the user experience can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of Internet of Things, and more specifically, to a device management method and system based on Internet of Things. Background Art

[0002] The Internet of Things (IoT) refers to a network system that connects various physical devices through the Internet to achieve information exchange and remote control between devices. With the development of the Internet of Things, more and more devices are connected to the Internet, including charging piles. Charging piles are the infrastructure for charging electric vehicles. They are usually installed in public places or private parking lots. In order to ensure the normal operation and safety of charging piles, it is crucial to effectively manage and monitor charging piles.

[0003] However, traditional charging pile management solutions mainly rely on manual inspections and fault alarms of charging piles, which is inefficient and requires a lot of human resources and time. Moreover, manual inspections are prone to missed inspections or misjudgments, and cannot monitor the status of charging piles in real time. In addition, traditional charging pile management solutions have a slow response speed. In other words, when a charging pile fails or is abnormal, it is necessary to wait for manual inspections or user reports to be notified. This will delay the time for troubleshooting and affect the user's charging experience.

[0004] Therefore, an optimized IoT-based device management solution is desired. Summary of the invention

[0005] In order to solve the above technical problems, this application is proposed. The embodiments of this application provide a device management method and system based on the Internet of Things. It can avoid the low efficiency and low accuracy problems caused by traditional manual inspections, thereby realizing remote management, fault diagnosis and predictive maintenance of charging piles, and further improving the reliability and user experience of charging piles.

[0006] According to one aspect of the present application, a device management method based on the Internet of Things is provided, which includes:

[0007] Acquire charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor;

[0008] Arranging the charging current values ​​at the plurality of predetermined time points into a charging current timing input vector according to the time dimension;

[0009] Performing current timing characteristic analysis on the charging current timing input vector to obtain charging current timing characteristics; and

[0010] Based on the charging current timing characteristics, it is determined whether there is any abnormality in the working state of the monitored charging pile.

[0011] According to another aspect of the present application, there is provided a device management system based on the Internet of Things, which includes:

[0012] A charging current value acquisition module, used to acquire the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor;

[0013] A vectorization module, used for arranging the charging current values ​​at the plurality of predetermined time points into a charging current timing input vector according to a time dimension;

[0014] a current timing characteristic analysis module, configured to perform current timing characteristic analysis on the charging current timing input vector to obtain a charging current timing characteristic; and

[0015] The working status analysis module is used to determine whether there is any abnormality in the working status of the monitored charging pile based on the charging current timing characteristics.

[0016] Compared with the prior art, the device management method and system based on the Internet of Things provided by the present application first obtains the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor, then arranges the charging current values ​​at the multiple predetermined time points into a charging current timing input vector according to the time dimension, then performs a current timing characteristic analysis on the charging current timing input vector to obtain a charging current timing characteristic, and finally, based on the charging current timing characteristic, determines whether the working state of the monitored charging pile is abnormal. In this way, the reliability of the charging pile and the user experience can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0018] Figure 1 The present invention is a flowchart of a device management method based on the Internet of Things according to an embodiment of the present application.

[0019] Figure 2 Schematic diagram of the architecture of the device management method based on the Internet of Things according to an embodiment of the present application.

[0020] Figure 3 This is a flowchart of sub-step S130 of the device management method based on the Internet of Things according to an embodiment of the present application.

[0021] Figure 4 This is a flowchart of sub-step S133 of the device management method based on the Internet of Things according to an embodiment of the present application.

[0022] Figure 5 The block diagram is a device management system based on the Internet of Things according to an embodiment of the present application.

[0023] Figure 6 This is an application scenario diagram of the device management method based on the Internet of Things according to an embodiment of the present application. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0025] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0026] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0027] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. Meanwhile, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0028] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0029] In response to the above-mentioned technical problems, the technical concept of the present application is to collect the charging current value of the charging pile in real time through a current sensor, and introduce data processing and analysis algorithms at the back end to perform timing analysis of the charging pile current, so as to monitor in real time whether the working status of the charging pile is abnormal. In this way, the low efficiency and low accuracy problems caused by traditional manual inspections can be avoided, thereby realizing remote management, fault diagnosis and predictive maintenance of the charging pile, and further improving the reliability of the charging pile and user experience.

[0030] Figure 1 The present invention is a flowchart of a device management method based on the Internet of Things according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of the device management method based on the Internet of Things according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the device management method based on the Internet of Things embodiment of the present application, the steps include: S110, obtaining the charging current value of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor; S120, arranging the charging current values ​​of the multiple predetermined time points according to the time dimension as a charging current timing input vector; S130, performing current timing characteristic analysis on the charging current timing input vector to obtain a charging current timing characteristic; and, S140, determining whether there is any abnormality in the working state of the monitored charging pile based on the charging current timing characteristic.

[0031] Specifically, in the technical solution of the present application, first, the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor are obtained. Next, considering that the charging current value of the charging pile has a dynamic change law of time series in the time dimension, that is, there is a time series correlation relationship between the charging current values ​​of the charging pile at the multiple predetermined time points. Therefore, in order to be able to capture and characterize the time series change pattern and trend of the charging current value of the charging pile, so as to accurately monitor the working status of the charging pile in real time, in the technical solution of the present application, it is necessary to arrange the charging current values ​​of the multiple predetermined time points according to the time dimension as a charging current time series input vector, so as to integrate the distribution information of the charging current value of the charging pile in the time series, and further capture and characterize the time series characteristics of the charging current value of the charging pile.

[0032] It should be understood that the charging current value of the charging pile has volatility and uncertainty in the time dimension, and it may have different change patterns and characteristics in different time periods in the timing. Therefore, in order to better capture the charging current timing characteristics of the charging pile, in the technical solution of the present application, the charging current timing input vector is further vector segmented to obtain a sequence of charging current local timing input vectors. By vector segmenting the charging current timing input vector, the entire timing sequence can be divided into multiple local timing input vectors, each local timing input vector represents the charging current change over a period of time. In this way, the charging current timing change characteristics of the charging pile can be more fully captured to more accurately determine whether the working status of the charging pile is normal.

[0033] Then, in order to improve the ability to capture subtle changes in the timing of the charging current value of the charging pile, in the technical solution of the present application, each charging current local timing input vector in the sequence of the charging current local timing input vector is further upsampled to increase the density and smoothness of the data parameters in timing, thereby facilitating a better subsequent representation of the working status characteristics of the charging pile.

[0034] Furthermore, considering that compared to a simple vector representation, a time series image can provide more information, including the time series variation pattern, fluctuation and trend of the current data. Therefore, after upsampling each charging current local time series input vector in the sequence of the charging current local time series input vector, the upsampled charging current local time series input vector is further converted from a vector to an image. Such a conversion can better capture the dynamic variation characteristics of the charging current time series of the charging pile. Based on this, in the technical solution of the present application, the sequence of the charging current local time series input vector is respectively passed through a format converter including an upsampling layer and a vector-image conversion layer to obtain a sequence of charging current local time series images.

[0035] Furthermore, considering that the charging current of the charging pile has a dynamic correlation characteristic of time series in the time dimension, in order to better capture the dynamic change characteristics of the charging current value in the time dimension, so as to more comprehensively and accurately represent the working state of the charging pile, in the technical solution of the present application, the sequence of the local time series images of the charging current is further subjected to feature mining in a charging current time series pattern feature extractor based on a three-dimensional convolutional neural network model, so as to extract the time series correlation feature distribution information between the local time series image features of the charging current in the time dimension, thereby obtaining a charging current time series feature diagram. In particular, here, by extracting the time series feature diagram of the charging current, the original charging current time series image data can be converted into a higher-level and more abstract representation, including the waveform, frequency, amplitude change and other feature information of the charging current, which can more accurately reflect the working state of the charging pile, and at the same time, it also helps to reduce the dimension and redundancy of the data, extract key feature information, and provide more useful input for subsequent charging pile state judgment and fault diagnosis.

[0036] It should be understood that in the charging current timing characteristic diagram, different spatial positions may contain different important information, and some positions contain more critical information about the timing changes of the charging current and the monitoring of the working status of the charging pile. Therefore, in the technical solution of the present application, the charging current timing characteristic diagram is further passed through the spatial attention layer to obtain a spatial dimension-manifested charging current timing characteristic diagram, so as to further highlight and strengthen the spatial dimension information in the feature diagram, so that the timing characteristics of the charging current are more significant and visual in space. By applying the spatial attention layer, the model can automatically learn and manifest the spatial dimension information in the charging current timing characteristic diagram, so that the key spatial positions in the feature diagram are more obvious and recognizable, which helps to further enhance the model's understanding and judgment of the working status of the charging pile, so that it can more accurately detect and identify abnormal or fault conditions.

[0037] Accordingly, if Figure 3 As shown, the current timing feature analysis is performed on the charging current timing input vector to obtain the charging current timing feature, including: S131, vector segmentation is performed on the charging current timing input vector to obtain a sequence of charging current local timing input vectors; S132, the sequence of charging current local timing input vectors is respectively passed through a format converter including an upsampling layer and a vector-image conversion layer to obtain a sequence of charging current local timing images; and, S133, feature extraction is performed on the sequence of charging current local timing images through a charging current timing pattern feature extractor based on a deep neural network model to obtain the charging current timing feature.

[0038] It should be understood that in step S131, the entire charging current time series input vector is divided into multiple local time series input vectors. The purpose of this is to divide the time series data of the charging current into smaller fragments for better subsequent processing and feature extraction. In step S132, the local time series input vector of the charging current is converted into a corresponding local time series image. This conversion process usually includes an upsampling layer and a vector-image conversion layer, which can convert the time series data into image data. This conversion can help the model better capture the features and patterns in the time series data. In step S133, a model based on a deep neural network, such as a convolutional neural network (CNN) or a recurrent neural network (RNN), is used to extract features from the local time series image of the charging current. These feature extractors can learn and capture key features in the image, such as shape, texture, and pattern. By extracting these features, a feature representation describing the charging current time series can be obtained for subsequent analysis and application. Combining the above three steps, by dividing the charging current time series data into local time series input vectors, then converting them into image form, and using a deep neural network model to extract features, a time series feature representation of the charging current can be obtained. These features can be used for applications such as pattern recognition, anomaly detection, and prediction of charging current.

[0039] Among them, in step S133, if Figure 4 As shown, feature extraction is performed on the sequence of local timing images of the charging current by a charging current timing pattern feature extractor based on a deep neural network model to obtain the charging current timing feature, including: S1331, passing the sequence of local timing images of the charging current through a charging current timing pattern feature extractor based on a three-dimensional convolutional neural network model to obtain a charging current timing feature graph; and, S1332, passing the charging current timing feature graph through a spatial attention layer to obtain a spatial dimension manifested charging current timing feature graph as the charging current timing feature.

[0040] It is worth mentioning that in step S1331, a model based on a three-dimensional convolutional neural network (3D CNN) is used to extract features from the local time series image of the charging current. 3D CNN is a variant of a convolutional neural network that can process image sequence data with a time dimension. By performing convolution operations in the time, space, and channel dimensions, 3D CNN can capture the time series patterns and features in the image sequence, so that the features extracted by the model can represent the time series features of the charging current. In step S1332, the charging current time series feature map is processed using a spatial attention layer to highlight specific spatial dimension features. The spatial attention layer can learn and weight features at different spatial positions so that important spatial dimension features are emphasized. In this way, the spatial dimension features in the charging current time series feature map can be visualized to make it more prominent and meaningful. The spatial dimension visualized charging current time series feature map obtained in this way can be used as the final charging current time series feature representation. Combining the above two steps, the charging current local time series image is feature extracted by a feature extractor based on a three-dimensional convolutional neural network model to obtain a charging current time series feature map. Then, the feature map is processed by the spatial attention layer to highlight the spatial dimension features, and the charging current timing feature map with spatial dimension explicitness is obtained as the final charging current timing feature representation. These features can better describe the timing pattern and characteristics of the charging current for further analysis and application.

[0041] Then, the spatial dimension manifested charging current time series characteristic diagram is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the working state of the monitored charging pile is abnormal. In other words, the dynamic characteristic distribution information of the charging current time series of the charging pile after spatial feature manifestation is used for classification processing, so as to monitor and judge whether the working state of the charging pile is abnormal in real time. In this way, the low efficiency and low accuracy caused by traditional manual inspection can be avoided, thereby realizing remote management, fault diagnosis and predictive maintenance of the charging pile, and further improving the reliability and user experience of the charging pile.

[0042] Accordingly, based on the charging current timing characteristics, determining whether there is an abnormality in the working state of the monitored charging pile includes: passing the spatial dimension manifested charging current timing characteristic diagram through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the working state of the monitored charging pile.

[0043] Specifically, the spatial dimension explicit charging current timing feature graph is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the working state of the monitored charging pile, including: expanding the spatial dimension explicit charging current timing feature graph into a spatial dimension explicit charging current timing feature vector according to a row vector or a column vector; using the fully connected layer of the classifier to fully connect encode the spatial dimension explicit charging current timing feature vector to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into the Softmax classification function of the classifier to obtain the classification result.

[0044] That is, in the technical solution of the present application, the labels of the classifier include the existence of an abnormality in the working state of the monitored charging pile (the first label), and the existence of an abnormality in the working state of the monitored charging pile (the second label), wherein the classifier determines which classification label the spatial dimension explicit charging current time series characteristic diagram belongs to by means of a soft maximum function. It is worth noting that the first label p1 and the second label p2 here do not contain artificially set concepts. In fact, during the training process, the computer model does not have the concept of "whether the working state of the monitored charging pile is abnormal". It only has two classification labels and the probability of the output feature under these two classification labels, that is, the sum of p1 and p2 is one. Therefore, the classification result of whether the working state of the monitored charging pile is abnormal is actually converted into a binary classification class probability distribution that conforms to the laws of nature through the classification label. In essence, the physical meaning of the natural probability distribution of the label is used, rather than the linguistic text meaning of "whether the working state of the monitored charging pile is abnormal".

[0045] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression and SVM are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification. However, this is prone to errors and is not efficient. Commonly used multi-classification methods include the Softmax classification function.

[0046] Further, in the technical solution of the present application, the device management method based on the Internet of Things also includes a training step: for training the format converter including the upsampling layer and the vector-image conversion layer, the charging current time series pattern feature extractor based on the three-dimensional convolutional neural network model, the spatial attention layer and the classifier. It should be understood that the training step is very important in the device management method based on the Internet of Things. Its role is to learn and extract the time series features of the charging current through the training model, and to build an effective classifier to classify the charging current or other tasks. Specifically, the role of the training step includes: 1. Learning feature representation: By training the format converter including the upsampling layer and the vector-image conversion layer, the charging current time series pattern feature extractor based on the three-dimensional convolutional neural network model and the spatial attention layer, the model can learn how to convert the time series data of the charging current into meaningful feature representations. These feature representations can capture patterns, trends and other important features in the charging current. 2. Improving classification performance: By training the classifier, the model can learn how to associate the extracted time series features with the category of the charging current or other tasks. The classifier can learn how to accurately classify charging current samples into different categories, such as normal and abnormal states, or different types of charging current patterns, through training data. By training the classifier, the performance and accuracy of the model in charging current classification or other tasks can be improved. 3. Generalization ability: Through the training step, the model can learn the general patterns and characteristics of the charging current, thereby improving its generalization ability. Generalization ability refers to the performance of the model on unseen data. Through training, the model can learn the general laws of charging current, rather than just remembering the specific samples in the training set. This enables the model to accurately classify and analyze new charging current data in practical applications. In summary, the role of the training step in the IoT-based device management method is to extract the timing characteristics of the charging current by learning and training the model, and build an effective classifier to classify the charging current or other tasks. This can improve the performance, accuracy, and generalization ability of the model, making it more reliable and effective in practical applications.

[0047] In one example, the training step includes: obtaining training data, the training data including training charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period, and a real value of whether the working state of the monitored charging pile is abnormal; arranging the training charging current values ​​at the multiple predetermined time points into a training charging current timing input vector according to the time dimension; vector segmenting the training charging current timing input vector to obtain a sequence of training charging current local timing input vectors; passing the sequence of training charging current local timing input vectors through the format converter comprising an upsampling layer and a vector-image conversion layer to obtain a sequence of training charging current local timing images; passing the sequence of training charging current local timing images through the charging current timing pattern feature based on the three-dimensional convolutional neural network model. extractor to obtain a training charging current timing feature graph; pass the training charging current timing feature graph through the spatial attention layer to obtain a training spatial dimension explicit charging current timing feature graph; pass the training spatial dimension explicit charging current timing feature graph through the classifier to obtain a classification loss function value; and, based on the classification loss function value and propagated in the direction of gradient descent, train the format converter including the upsampling layer and the vector-image conversion layer, the charging current timing pattern feature extractor based on the three-dimensional convolutional neural network model, the spatial attention layer and the classifier, and correct the training spatial dimension explicit charging current timing feature vector after each training spatial dimension explicit charging current timing feature graph is expanded and classified iteratively through the classifier.

[0048] In one example, the training space dimension explicit charging current timing characteristic diagram is passed through the classifier to obtain a classification loss function value, including: the classifier processes the training space dimension explicit charging current timing characteristic diagram with the following training classification formula to generate a training classification result; wherein the training classification formula is:

[0049] softmax{(W n ,B n ):…:(W 1 ,B 1 )|V}

[0050] Wherein, V represents the training space dimension explicit charging current time series feature vector obtained by expanding the training space dimension explicit charging current time series feature graph, and W 1 To W n is the weight matrix of each fully connected layer, B 1 To B nRepresenting the bias matrix of each fully connected layer; and calculating the cross entropy value between the training classification result and the true value as the classification loss function value.

[0051] In particular, in the technical solution of the present application, the charging current time series input vector is vector segmented to obtain a sequence of charging current local time series input vectors, the sequence of charging current local time series input vectors is passed through a format converter including an upsampling layer and a vector-to-image conversion layer to obtain a sequence of charging current local time series images, and the sequence of charging current local time series images is passed through a charging current time series pattern feature extractor based on a three-dimensional convolutional neural network model. In this way, the time series correlation features of the charging current value in the local time domain based on vector segmentation in the global time domain and in the multi-scale time domain subspace of the subdivided time domain in the local time domain based on the vector-to-image format conversion can be extracted. In this way, the spatial attention layer can further strengthen the distribution of timing correlation features in local time domain subspaces under certain local time domains. However, if the spatial dimension explicit charging current timing feature map is taken as a whole, the timing correlation feature expression under each time domain subspace may be unbalanced, which makes the spatial dimension explicit charging current timing feature map have rich geometric shape details in the high-dimensional feature space, resulting in category representation distortion when classified by the classifier to map to the class probability domain, that is, affecting the category representation of the spatial dimension explicit charging current timing feature map for the category predetermined by the classifier, thereby reducing the accuracy of the classification result.

[0052] Based on this, each time the training space dimension explicit charging current time series feature vector after the training space dimension explicit charging current time series feature graph is expanded is classified and iterated by the classifier, the training space dimension explicit charging current time series feature vector is corrected, including: correcting the training space dimension explicit charging current time series feature vector by the following correction formula to obtain a corrected training space dimension explicit charging current time series feature vector; wherein the correction formula is:

[0053]

[0054] Wherein, V is the training space dimension explicit charging current time series feature vector, p is the class probability value obtained by the classifier of the training space dimension explicit charging current time series feature vector V, and v i is the i-th eigenvalue of the charging current time series eigenvector V in the training space dimension, is the mean of all eigenvalues ​​of the charging current time series eigenvector V in the training space dimension, ‖V‖ 1 represents the norm of the training space dimension manifesting the charging current time series feature vector V, and α is the weight hyperparameter, vi ′ is the i-th eigenvalue of the charging current time series eigenvector manifested in the training space dimension after correction.

[0055] Specifically, while maintaining the geometric shape details of the high-dimensional feature manifold under the class probability mapping of the training space dimension explicit charging current time series feature vector V, the shape attribute editing is performed on the feature value details relative to the variational representation of the feature set of the training space dimension explicit charging current time series feature vector V through the distortion consultation of the class probability, thereby mapping the class probability distortion of the training space dimension explicit charging current time series feature vector V to the latent class space feature representation, and then consulting fusion is performed by supplementing the basic low-rank constrained latent inversion representation to bridge the gap between the edited consulting representation and the original geometric shape details, thereby improving the class representation of the high-dimensional features of the training space dimension explicit charging current time series feature vector V, thereby improving the accuracy of the classification results. In this way, it is possible to automatically monitor whether the working status of the charging pile is abnormal in real time, thereby realizing remote management, fault diagnosis and predictive maintenance of the charging pile, and further improving the reliability and user experience of the charging pile.

[0056] In summary, the IoT-based device management method according to the embodiment of the present application is explained, which can improve the reliability and user experience of charging piles.

[0057] Figure 5 FIG. 1 is a block diagram of an IoT-based device management system 100 according to an embodiment of the present application. Figure 5 As shown, according to the IoT-based device management system 100 of the embodiment of the present application, it includes: a charging current value acquisition module 110, which is used to obtain the charging current value of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by a current sensor; a vectorization module 120, which is used to arrange the charging current values ​​of the multiple predetermined time points according to the time dimension as a charging current timing input vector; a current timing feature analysis module 130, which is used to perform current timing feature analysis on the charging current timing input vector to obtain a charging current timing feature; and a working status analysis module 140, which is used to determine whether there is any abnormality in the working status of the monitored charging pile based on the charging current timing feature.

[0058] In one example, in the above-mentioned IoT-based device management system 100, the current timing feature analysis module 130 includes: a vector segmentation unit, which is used to perform vector segmentation on the charging current timing input vector to obtain a sequence of charging current local timing input vectors; a format conversion unit, which is used to pass the sequence of charging current local timing input vectors through a format converter including an upsampling layer and a vector-image conversion layer to obtain a sequence of charging current local timing images; and a feature extraction unit, which is used to perform feature extraction on the sequence of charging current local timing images through a charging current timing pattern feature extractor based on a deep neural network model to obtain the charging current timing feature.

[0059] Here, those skilled in the art can understand that the specific functions and operations of each module in the above-mentioned IoT-based device management system 100 have been described in detail above. Figures 1 to 4 The description of the IoT-based device management method has been described in detail, and therefore, its repeated description will be omitted.

[0060] As described above, the device management system 100 based on the Internet of Things according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with a device management algorithm based on the Internet of Things. In one example, the device management system 100 based on the Internet of Things according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the device management system 100 based on the Internet of Things can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the device management system 100 based on the Internet of Things can also be one of the many hardware modules of the wireless terminal.

[0061] Alternatively, in another example, the IoT-based device management system 100 and the wireless terminal may also be separate devices, and the IoT-based device management system 100 may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0062] Figure 6 FIG. 1 is an application scenario diagram of the device management method based on the Internet of Things according to an embodiment of the present application. Figure 6 As shown, in this application scenario, first, the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor are obtained (for example, Figure 6 D), and then, the charging current values ​​at the plurality of predetermined time points are input to a server (for example, Figure 6In S) shown in , the server is able to use the IoT-based device management algorithm to process the charging current values ​​at the multiple predetermined time points to obtain a classification result indicating whether there is any abnormality in the working status of the monitored charging pile.

[0063] Furthermore, the present application designs corresponding software based on the device management method and system based on the Internet of Things, which includes a data acquisition module, an upload module, a data processing and analysis module, an early warning module, an export module for inspection and early warning results, a kanban module, and a management background. Among them, the data acquisition module mainly collects the operating conditions of the installed charging piles, such as whether there are broken ones, those that do not meet the operating standards, etc. The upload module supports the uploading of entity forms, on-site photos, punch-in photos, etc., and the uploaded forms are automatically entered into the system. When entering the system, in addition to the form of the equipment operation status, form fields are also provided to facilitate internal management. The data processing and analysis module determines whether there is an equipment abnormality based on the data obtained through collection and inspection. The early warning module summarizes and feedbacks the early warning information based on the abnormal information. The export module for inspection and early warning results can automatically export the inspection and early warning results, and automatically generate reports for use by the supervision, management and law enforcement departments. The kanban module, in this kanban, can view the conclusions summarized and analyzed by multiple modules in real time and intuitively.

[0064] In addition, it will be appreciated by those skilled in the art that various aspects of the present application may be illustrated and described by a number of patentable categories or situations, including any new and useful process, machine, product or combination of substances, or any new and useful improvements thereto. Accordingly, various aspects of the present application may be performed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The above hardware or software may all be referred to as "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of the present application may be represented as a computer product located in one or more computer-readable media, which includes computer-readable program code.

[0065] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the invention belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0066] The above is an explanation of the present invention and should not be considered as a limitation thereof. Although several exemplary embodiments of the present invention have been described, it will be readily appreciated by those skilled in the art that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present invention. Therefore, all such modifications are intended to be included within the scope of the present invention as defined in the claims. It should be understood that the above is an explanation of the present invention and should not be considered as being limited to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present invention is defined by the claims and their equivalents.

Claims

1. A device management method based on the Internet of Things, characterized in that: include: Acquire charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor; Arranging the charging current values ​​at the plurality of predetermined time points into a charging current timing input vector according to the time dimension; Performing current timing characteristic analysis on the charging current timing input vector to obtain charging current timing characteristics; as well as Based on the charging current timing characteristics, determining whether the working state of the monitored charging pile is abnormal; Performing current timing characteristic analysis on the charging current timing input vector to obtain the charging current timing characteristic includes: Performing vector segmentation on the charging current timing input vector to obtain a sequence of charging current local timing input vectors; Passing the sequence of charging current local time series input vectors through a format converter including an upsampling layer and a vector-to-image conversion layer to obtain a sequence of charging current local time series images; and Extracting features from a sequence of local time series images of the charging current by a charging current time series pattern feature extractor based on a deep neural network model to obtain the charging current time series features; The method further comprises: extracting features of the sequence of the charging current local time series images by a charging current time series pattern feature extractor based on a deep neural network model to obtain the charging current time series features, including: Passing the sequence of the charging current local time-series images through a charging current time-series pattern feature extractor based on a three-dimensional convolutional neural network model to obtain a charging current time-series feature graph; and Passing the charging current timing feature graph through a spatial attention layer to obtain a spatial dimension explicit charging current timing feature graph as the charging current timing feature; Based on the charging current timing characteristics, determining whether the working state of the monitored charging pile is abnormal includes: The spatial dimension manifested charging current time series characteristic diagram is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the working state of the monitored charging pile; The method further comprises a training step for training the format converter comprising an upsampling layer and a vector-to-image conversion layer, the charging current timing pattern feature extractor based on a three-dimensional convolutional neural network model, the spatial attention layer and the classifier; The training step comprises: Acquire training data, the training data including training charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period, and a real value of whether the working state of the monitored charging pile is abnormal; Arranging the training charging current values ​​at the plurality of predetermined time points into a training charging current timing input vector according to the time dimension; Performing vector segmentation on the training charging current timing input vector to obtain a sequence of training charging current local timing input vectors; Passing the sequence of the training charging current local time series input vectors through the format converter comprising the upsampling layer and the vector-to-image conversion layer respectively to obtain a sequence of training charging current local time series images; Passing the sequence of the training charging current local time series images through the charging current time series pattern feature extractor based on the three-dimensional convolutional neural network model to obtain a training charging current time series feature graph; Passing the training charging current time series feature graph through the spatial attention layer to obtain a training spatial dimension explicit charging current time series feature graph; Passing the training space dimension manifested charging current time series characteristic diagram through the classifier to obtain a classification loss function value; and The format converter including the upsampling layer and the vector-to-image conversion layer, the charging current timing pattern feature extractor based on the three-dimensional convolutional neural network model, the spatial attention layer and the classifier are trained based on the classification loss function value and propagated in the direction of gradient descent, and the training space-dimensional explicit charging current timing feature vector is corrected each time the training space-dimensional explicit charging current timing feature vector after the training space-dimensional explicit charging current timing feature graph is expanded passes through the classifier for classification iteration; Each time the training space dimension explicit charging current time series feature vector after the training space dimension explicit charging current time series feature graph is expanded is classified and iterated by the classifier, the training space dimension explicit charging current time series feature vector is corrected, including: correcting the training space dimension explicit charging current time series feature vector by the following correction formula to obtain a corrected training space dimension explicit charging current time series feature vector; wherein the correction formula is: in, is the characteristic vector of the charging current timing in the training space dimension, is the training space dimension manifesting the charging current time series feature vector The class probability value obtained by the classifier, is the training space dimension manifesting the charging current time series feature vector No. eigenvalues, is the training space dimension manifesting the charging current time series feature vector The mean of all eigenvalues ​​of Represents the training space dimension manifesting the charging current time series feature vector The one-norm of is the weight hyperparameter, is the first dimension of the charging current time series feature vector after correction training space dimension feature values.

2. The device management method based on the Internet of Things according to claim 1 is characterized in that: Passing the training space dimension manifested charging current time series characteristic diagram through the classifier to obtain a classification loss function value, including: The classifier processes the training space dimension explicit charging current timing characteristic diagram using the following training classification formula to generate a training classification result; wherein the training classification formula is: in, represents the training space dimension explicit charging current time series feature vector obtained by expanding the training space dimension explicit charging current time series feature graph, to is the weight matrix of each fully connected layer, to represents the bias matrix of each fully connected layer; and A cross entropy value between the training classification result and the true value is calculated as the classification loss function value.

3. An IoT-based device management system for executing the IoT-based device management method according to claim 1, characterized in that: include: A charging current value acquisition module, used to acquire the charging current values ​​of the monitored charging pile at multiple predetermined time points within a predetermined time period collected by the current sensor; A vectorization module, used for arranging the charging current values ​​at the plurality of predetermined time points into a charging current timing input vector according to a time dimension; A current timing characteristic analysis module, used for performing current timing characteristic analysis on the charging current timing input vector to obtain a charging current timing characteristic; as well as The working status analysis module is used to determine whether there is any abnormality in the working status of the monitored charging pile based on the charging current timing characteristics.

4. The device management system based on the Internet of Things according to claim 3 is characterized in that: The current timing characteristic analysis module comprises: A vector segmentation unit, used for performing vector segmentation on the charging current timing input vector to obtain a sequence of charging current local timing input vectors; a format conversion unit, configured to pass the sequence of charging current local time series input vectors through a format converter including an upsampling layer and a vector-image conversion layer to obtain a sequence of charging current local time series images; and A feature extraction unit is used to extract features from the sequence of local time series images of the charging current through a charging current time series pattern feature extractor based on a deep neural network model to obtain the charging current time series features.

Citation Information

Patent Citations

  • Network traffic identification method and system

    CN117113262A

  • Electric vehicle charging pile fault early warning system and method based on big data

    CN117818414A