Monitoring system and method for stranding machine

Abstract

The present application discloses a monitoring system and method for a stranding machine, which collects various data parameters of the stranding machine in real time during operation by deploying various data monitoring modules on the stranding machine, such as a speed monitoring module, a temperature monitoring module, and a current detection module, and uses a fault diagnosis module to perform a time-series collaborative analysis of these stranding machine operation data using data processing and analysis algorithms, so as to determine whether the stranding machine's operating status is normal, and to promptly issue a fault signal and display it when an abnormality is detected, so that corresponding measures can be taken to repair it later and avoid production interruptions and losses. In this way, real-time monitoring and automatic control of the stranding machine's operating status can be achieved, thereby ensuring the quality of the stranded wire, improving production efficiency, and reducing failures and losses.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to a monitoring system and method for a stranding machine. Background Art

[0002] A stranding machine is a device that twists multiple metal or fiber wires into a composite wire according to specific rules. It is widely used in industries such as electrical cables, optical cables, and wire ropes. Various parameters in the stranding process, such as stranding speed, tension, temperature, and current, directly impact the quality and performance of the resulting stranded wire. Therefore, real-time monitoring and control of the stranding machine is crucial for improving stranding quality and production efficiency. However, traditional stranding machine monitoring systems typically employ rule-based methods, requiring manual setting of thresholds and alarm conditions for various parameters. These systems lack intelligence and adaptability, making them ineffective in handling complex and changing stranding scenarios.

[0003] Therefore, an optimized monitoring system for a stranding machine is desired. Summary of the Invention

[0004] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a monitoring system and method for a stranding machine, which collects various data parameters during the operation of the stranding machine in real time through various data monitoring modules deployed in the stranding machine, such as a speed monitoring module, a temperature monitoring module and a current detection module, and uses a fault diagnosis module to perform a time series collaborative analysis of these stranding machine operation data using data processing and analysis algorithms, so as to determine whether the operation status of the stranding machine is normal, and promptly send out a fault signal and display it when an abnormality is detected, so as to facilitate the subsequent adoption of corresponding measures for repair, thereby avoiding the occurrence of production interruptions and losses. In this way, real-time monitoring and automatic control of the operation status of the stranding machine can be achieved, thereby ensuring the quality of the stranded wire, improving production efficiency, and reducing failures and losses.

[0005] According to one aspect of the present application, a monitoring system for a stranding machine is provided, comprising:

[0006] Speed ​​monitoring module, used to measure the speed and tension of the monitored stranding machine, and synchronously control the speed ratio of each axis;

[0007] A temperature monitoring module, used to measure the temperature of the monitored stranding machine and adjust the switch and air volume of the cooling fan;

[0008] A current monitoring module, used to measure the current of the monitored stranding machine and determine the working state and load condition of the stranding machine;

[0009] a fault diagnosis module, configured to detect a fault signal of the monitored stranding machine and display the fault signal;

[0010] The data recording module is used to record the operating data of the monitored stranding machine and generate stranding reports and statistical analysis.

[0011] According to another aspect of the present application, a monitoring method for a stranding machine is provided, comprising:

[0012] Measure the speed and tension of the monitored stranding machine and synchronously control the speed ratio of each axis;

[0013] Measuring the temperature of the monitored stranding machine and adjusting the switch and air volume of the cooling fan;

[0014] Measuring the current of the monitored stranding machine and determining the working state and load condition of the stranding machine;

[0015] detecting a fault signal of the monitored stranding machine and displaying the fault signal;

[0016] Record the operating data of the monitored stranding machine, and generate stranding reports and statistical analysis.

[0017] Compared with the existing technology, the present application provides a monitoring system and method for a stranding machine, which collects various data parameters during the operation of the stranding machine in real time by deploying various data monitoring modules on the stranding machine, such as a speed monitoring module, a temperature monitoring module, and a current detection module, and uses a fault diagnosis module to perform a time-series collaborative analysis of these stranding machine operation data using data processing and analysis algorithms, so as to determine whether the operation status of the stranding machine is normal, and to promptly send out a fault signal and display it when an abnormality is detected, so that corresponding measures can be taken to repair it later, avoiding production interruptions and losses. In this way, real-time monitoring and automatic control of the operation status of the stranding machine can be achieved, thereby ensuring the quality of the stranded wire, improving production efficiency, and reducing failures and losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0019] Figure 1 is a block diagram of a monitoring system for a stranding machine according to an embodiment of the present application;

[0020] Figure 2 1 is a system architecture diagram of a monitoring system for a stranding machine according to an embodiment of the present application;

[0021] Figure 3 is a block diagram of a training phase of a monitoring system for a stranding machine according to an embodiment of the present application;

[0022] Figure 4 is a block diagram of a fault diagnosis module in a monitoring system for a stranding machine according to an embodiment of the present application;

[0023] Figure 5 4 is a block diagram of a data pre-processing and aggregation unit in a monitoring system for a stranding machine according to an embodiment of the present application;

[0024] Figure 6 1 is a block diagram of a fault signal detection unit in a monitoring system of a stranding machine according to an embodiment of the present application;

[0025] Figure 7 4 is a flow chart of a monitoring method for a stranding machine according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] 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 herein.

[0027] As used in this application and the claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not intended to refer to the singular but may include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0028] Although the present application makes various references to certain modules in the system according to 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 illustrative only, and different aspects of the system and method can use different modules.

[0029] Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the various steps may be processed in reverse order or simultaneously, as needed. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0030] 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 herein.

[0031] Traditional wire stranding machine monitoring systems typically use rule-based methods, requiring manual setting of thresholds and alarm conditions for various parameters. These systems lack intelligence and adaptability, and are unable to effectively handle complex and changing wire stranding scenarios. Therefore, an optimized wire stranding machine monitoring system is desired.

[0032] In the technical solution of the present application, a monitoring system for a stranding machine is proposed. Figure 1 4 is a block diagram of a monitoring system for a stranding machine according to an embodiment of the present application. Figure 2 FIG. 1 is a system architecture diagram of a monitoring system for a stranding machine according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the monitoring system 300 of the stranding machine according to the embodiment of the present application includes: a speed monitoring module 310, which is used to measure the speed and tension of the monitored stranding machine, and synchronously control the speed ratio of each axis; a temperature monitoring module 320, which is used to measure the temperature of the monitored stranding machine, and adjust the switch and air volume of the cooling fan; a current monitoring module 330, which is used to measure the current of the monitored stranding machine, and determine the working status and load condition of the stranding machine; a fault diagnosis module 340, which is used to detect the fault signal of the monitored stranding machine, and display the fault signal; a data recording module 350, which is used to record the operating data of the monitored stranding machine, and generate stranding reports and statistical analysis.

[0033] Specifically, the speed monitoring module 310 is used to measure the speed and tension of the monitored stranding machine and synchronously control the speed ratio of each axis. It will be appreciated that speed measurement provides accurate monitoring of the stranding machine's operating status to ensure it operates within a predetermined speed range. Tension measurement determines whether the stranding machine is operating within the appropriate tension range to avoid excessively tight or loose strands, which could affect stranding quality or cause safety issues. By synchronously controlling the speed ratios of each axis, the motion of each axis of the stranding machine can be coordinated and consistent, thereby achieving high-quality stranding operations.

[0034] Specifically, the temperature monitoring module 320 is used to measure the temperature of the monitored stranding machine and adjust the cooling fan's on / off and air volume. It will be appreciated that temperature measurement can help monitor heat accumulation in the stranding machine and detect overheating. By monitoring temperature, timely measures can be taken to prevent damage to the stranding machine or other safety issues caused by overheating.

[0035] Specifically, the current monitoring module 330 is used to measure the current of the monitored stranding machine and determine the stranding machine's operating status and load. It should be understood that by monitoring the stranding machine's current, its current operating status can be determined. Different operating states typically correspond to different current levels; abnormal current fluctuations or current exceeding a preset threshold may indicate a stranding machine fault or abnormality. By monitoring the current, these abnormalities can be detected promptly, triggering appropriate alarms or shutdown protection mechanisms.

[0036] In particular, the fault diagnosis module 340 is used to detect the fault signal of the monitored stranding machine and display the fault signal. In particular, in a specific example of the present application, as Figure 4 As shown, the fault diagnosis module 340 includes: a data acquisition unit 341, which is used to obtain the stranding speed value, tension value, temperature value and current value at multiple predetermined time points within a predetermined time period collected by each monitoring module deployed on the monitored stranding machine; a data parameter timing arrangement unit 342, which is used to arrange the stranding speed value, tension value, temperature value and current value at the multiple predetermined time points into a stranding speed timing input vector, a tension timing input vector, a temperature timing input vector and a current timing input vector according to the time dimension; a data preprocessing aggregation unit 343, which is used to process the stranding speed timing input vector, the tension timing input vector, the temperature timing input vector and the current timing input vector. The sequential input vector is converted into the image domain and the parameter multi-channel aggregation is performed to obtain a parameter multi-channel time series image; a data time series image local feature extraction unit 344 is used to extract features of the parameter multi-channel time series image through an image area feature extractor based on a deep neural network model to obtain a sequence of parameter multi-channel time series association semantic feature vectors; a data time series image local semantic association encoding unit 345 is used to perform context association encoding on the sequence of parameter multi-channel time series association semantic feature vectors to obtain a multi-parameter global time series context semantic association feature; a fault signal detection unit 346 is used to determine whether to issue the fault signal based on the multi-parameter global time series context semantic association feature.

[0037] Specifically, the data acquisition unit 341 is used to obtain the stranding speed value, tension value, temperature value and current value at multiple predetermined time points within a predetermined time period, which are collected by the various monitoring modules deployed on the monitored stranding machine. In one example, the various monitoring modules may be composed of a speed sensor, a tension sensor, a temperature sensor and a current sensor. It is worth mentioning that a speed sensor is a device or sensor for measuring the speed of an object. It can detect the motion state of an object and output a corresponding speed signal, usually provided in the form of an electrical signal. A tension sensor is a sensor for measuring the tension or tension in an object or system. It can detect the tension or tension applied to an object and convert it into a corresponding electrical signal or mechanical signal output. A temperature sensor is a device or sensor for measuring the temperature of an environment or an object. It can detect the heat of the surrounding environment or an object and convert it into a corresponding electrical signal output. A current sensor is a device or sensor for measuring current. It can detect the current in a circuit and convert it into a corresponding electrical signal output.

[0038] Specifically, the data parameter time series arrangement unit 342 is used to arrange the stranding speed values, tension values, temperature values, and current values ​​at the multiple predetermined time points into a stranding speed time series input vector, a tension time series input vector, a temperature time series input vector, and a current time series input vector, respectively, according to the time dimension. It should be understood that considering that the stranding speed values, tension values, temperature values, and current values ​​of the monitored stranding machine not only have a time series dynamic change law in the time dimension, but also have a time series collaborative correlation relationship between the monitoring parameters of these stranding machines, this correlation relationship is of great significance for the detection and judgment of the operating status of the stranding machine. Therefore, in the technical solution of the present application, it is necessary to arrange the stranding speed values, tension values, temperature values, and current values ​​at the multiple predetermined time points into a stranding speed time series input vector, a tension time series input vector, a temperature time series input vector, and a current time series input vector, respectively, according to the time dimension, so as to respectively integrate the time series distribution information of the stranding speed values, tension values, temperature values, and current values ​​of the monitored stranding machine.

[0039] Specifically, the data pre-processing aggregation unit 343 is used to perform image domain conversion and parameter multi-channel aggregation processing on the strand speed time series input vector, the tension time series input vector, the temperature time series input vector, and the current time series input vector to obtain a parameter multi-channel time series image. In particular, in a specific example of the present application, if Figure 5As shown, the data preprocessing aggregation unit 343 includes: a vector-image domain conversion subunit 3431, which is used to pass the strand speed timing input vector, the tension timing input vector, the temperature timing input vector and the current timing input vector through a vector-image converter to obtain a strand speed timing image, a tension timing image, a temperature timing image and a current timing image; an image channel aggregation subunit 3432, which is used to aggregate the strand speed timing image, the tension timing image, the temperature timing image and the current timing image along the channel dimension to obtain the parameter multi-channel timing image.

[0040] More specifically, the vector-to-image domain conversion subunit 3431 is configured to pass the strand speed time-series input vector, the tension time-series input vector, the temperature time-series input vector, and the current time-series input vector through a vector-to-image converter to obtain a strand speed time-series image, a tension time-series image, a temperature time-series image, and a current time-series image, respectively. Considering that time-series image data can provide more information than a simple time-series vector representation, including the time-series relationship, fluctuations, and trend changes of strand speed, tension, temperature, and current data, in the technical solution of the present application, the strand speed time-series input vector, the tension time-series input vector, the temperature time-series input vector, and the current time-series input vector are further passed through a vector-to-image converter to obtain a strand speed time-series image, a tension time-series image, a temperature time-series image, and a current time-series image, respectively. In this way, the time-series information of the parameters can be converted into spatial information, enabling the deep learning algorithm to better capture the time-series correlations and change trends between the parameters. In other words, the vector-to-image converter can convert the stranding speed time-series input vector, the tension time-series input vector, the temperature time-series input vector, and the current time-series input vector into corresponding time-series images. These time-series images can contain information such as parameter variation trends, periodic fluctuations, and abnormal peaks, more intuitively reflecting the state and characteristics of the stranding process.

[0041] More specifically, the image channel aggregation subunit 3432 is configured to aggregate the stranding speed time-series images, the tension time-series images, the temperature time-series images, and the current time-series images along the channel dimension to generate the parameter multi-channel time-series image. To integrate the time-series information of different parameters for better analysis and correlation of the stranding process, the stranding speed time-series images, the tension time-series images, the temperature time-series images, and the current time-series images are further aggregated along the channel dimension to generate the parameter multi-channel time-series image.

[0042] It should be understood that during the stranding process, parameters such as stranding speed, tension, temperature, and current are often interrelated, and their changing trends and abnormal conditions may affect each other. By aggregating the time series images of these parameters along the channel dimension, their time series information can be integrated together to form a multi-channel time series image. Here, the multi-channel time series image of parameters can provide more comprehensive information, including the time series change trends, fluctuations, abnormal peaks, etc. between different parameters. Such an image can better reflect the state and characteristics of the stranding process, allowing deep learning algorithms to more accurately capture the correlation and change patterns between parameters.

[0043] It is worth mentioning that in other specific examples of the present application, the strand speed timing input vector, the tension timing input vector, the temperature timing input vector and the current timing input vector can also be subjected to image domain conversion and parameter multi-channel aggregation processing to obtain a parameter multi-channel timing image by other means, for example: converting each timing input vector into the image domain; performing parameter multi-channel aggregation processing on the converted image; superimposing the channels of each image together to form a multi-channel image. Each channel represents the information of an input vector; performing an average or weighted average operation on the corresponding pixels of each image to obtain an aggregated image. The pixel value of each channel represents the average or weighted average of the corresponding input vector; using computer vision or image processing technology to extract features from each image, and then using these features as channels to form a multi-channel image; generating a parameter multi-channel timing image based on the image after aggregation processing.

[0044] Specifically, the data time series image local feature extraction unit 344 is used to perform feature extraction on the parameter multi-channel time series image through an image region feature extractor based on a deep neural network model to obtain a sequence of parameter multi-channel time series associated semantic feature vectors. It should be understood that in the wire stranding machine monitoring system, the parameter multi-channel time series image may contain multiple key areas, each area corresponding to different parameter information. Considering that the RCNN model can be used for target detection and regional feature extraction in images, it can identify different areas in the image and extract feature information of these areas. Therefore, in order to further extract and represent the key features in the image, in the technical solution of the present application, the parameter multi-channel time series image is passed through an image region feature extractor based on the RCNN model to obtain a sequence of parameter multi-channel time series associated semantic feature vectors.

[0045] It is worth noting that RCNN (Region-based Convolutional Neural Network) is an object detection model that is one of the important milestones in the field of object detection. The core idea of ​​the RCNN model is to transform the object detection task into a candidate region classification problem.

[0046] Accordingly, in one possible implementation, the parameter multi-channel time series image can be passed through an image region feature extractor based on an RCNN model to obtain a sequence of parameter multi-channel time series associated semantic feature vectors through the following steps: for example, selecting an image region feature extractor based on an RCNN model, such as Faster R-CNN, Mask R-CNN, etc.; training the selected RCNN model; extracting image region features from each time series image frame in the parameter multi-channel time series image using the trained RCNN model; associating the extracted image region feature vectors in time series to form a sequence of parameter multi-channel time series associated semantic feature vectors. Time series association can be achieved using a target tracking algorithm or a simple time window sliding method; and performing appropriate feature representation on the sequence of parameter multi-channel time series associated semantic feature vectors.

[0047] Specifically, the data temporal image local semantic association encoding unit 345 is used to perform context association encoding on the sequence of the parameter multi-channel temporal association semantic feature vectors to obtain a multi-parameter global temporal context semantic association feature. Considering that in the process of detecting the state of the stranding machine, the multi-parameter temporal local feature information contained in each area of ​​the parameter multi-channel temporal image has global context association information, which includes long-term dependence, dynamic change trends and mutual influence between parameters. Therefore, in order to be able to use this multi-parameter global temporal context semantic association feature to detect the operating state of the stranding machine, it is obvious that the accuracy of anomaly detection can be improved. Based on this, in the technical solution of the present application, the sequence of the parameter multi-channel temporal association semantic feature vectors is further passed through a converter-based local feature context association encoder to obtain a multi-parameter global temporal context semantic association feature vector. In this way, the global temporal association features between the parameters of the stranding machine, including long-term dependence, dynamic change trends and mutual influence, can be better captured to facilitate subsequent analysis and detection tasks, such as anomaly detection, prediction and optimization. More specifically, the sequence of the parameter multi-channel temporal association semantic feature vectors is passed through a converter-based local feature context association encoder to obtain a multi-parameter global temporal context semantic association feature vector as the multi-parameter global temporal context semantic association feature, including: arranging the sequence of the parameter multi-channel temporal association semantic feature vectors in one dimension to obtain a global parameter multi-channel temporal association semantic feature vector; calculating the product between the global parameter multi-channel temporal association semantic feature vector and the transposed vector of each parameter multi-channel temporal association semantic feature vector in the sequence of the parameter multi-channel temporal association semantic feature vector to obtain multiple self-attention association matrices; and performing a multi-dimensional multi-processor multi-processor multi-processor multi-attention association on the multiple self-attention association matrices respectively. Each self-attention association matrix in the connection matrix is ​​normalized to obtain a plurality of standardized self-attention association matrices; each standardized self-attention association matrix in the plurality of standardized self-attention association matrices is subjected to a Softmax classification function to obtain a plurality of probability values; each parameter multi-channel temporal association semantic feature vector in the sequence of the parameter multi-channel temporal association semantic feature vector is weighted using each probability value in the plurality of probability values ​​to obtain the plurality of context semantic parameter multi-channel temporal association feature vectors; and the plurality of context semantic parameter multi-channel temporal association feature vectors are cascaded to obtain the multi-parameter global temporal context semantic association feature vector.

[0048] Specifically, the fault signal detection unit 346 is used to determine whether to issue the fault signal based on the multi-parameter global temporal context semantic association feature. In particular, in a specific example of the present application, Figure 6As shown, the fault signal detection unit 346 includes: a stranding machine operating status detection subunit 3461, which is used to pass the multi-parameter global temporal context semantic association feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the operating status of the monitored stranding machine; a fault signal issuance judgment subunit 3462, which is used to determine whether to issue the fault signal based on the classification result.

[0049] More specifically, the stranding machine operating status detection subunit 3461 is used to pass the multi-parameter global temporal context semantic association feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether there is an abnormality in the operating status of the monitored stranding machine. That is, the multi-parameter global temporal context semantic association feature information of the monitored stranding machine is used for classification processing to perform classification detection on whether the operating status of the stranding machine is abnormal. And based on the classification result, it is determined whether the fault signal is issued. In this way, it is possible to automatically determine whether the operating status of the stranding machine is normal, and when an abnormality is detected, a fault signal is issued and displayed in time, so that corresponding measures can be taken later to repair it and avoid production interruption and loss. In a specific example, the multi-parameter global temporal context semantic association feature vector 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 operating status of the monitored stranding machine, including: using multiple fully connected layers of the classifier to fully connect encode the multi-parameter global temporal context semantic association feature vector to obtain an encoded classification feature vector; and passing the encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0050] More specifically, the fault signal issuance determination subunit 3462 is configured to determine whether to issue the fault signal based on the classification results. In other words, in one example, based on the fault determination result obtained from the classification results, if a fault is determined to exist, a corresponding fault signal is issued. This can take the form of an alarm, notification, or report, allowing for timely implementation of appropriate maintenance or remedial measures.

[0051] It is worth mentioning that in other specific examples of the present application, it is also possible to determine whether to issue the fault signal based on the multi-parameter global time series context semantic association features in other ways, such as extracting features from the multi-parameter global time series data. This may include pre-processing the time series data of each parameter and then extracting features;

[0052] A context model is built to capture the temporal relationships between parameters. The extracted features are associated with the context model to obtain global temporal context semantic association features. The trained fault determination model is used to determine whether to issue a fault signal. Based on the fault determination results, if a fault is confirmed, a corresponding fault signal is issued. This can take the form of an alarm, notification, or report, allowing timely maintenance or remediation measures to be taken.

[0053] Specifically, the data recording module 350 is used to record the operating data of the monitored stranding machine, as well as generate stranding reports and statistical analysis. In other words, by recording the stranding machine's operating data, detailed operating information such as speed, tension, temperature, and current can be obtained. This data can be used for subsequent analysis and evaluation to understand changes in the stranding machine's performance and operating status. Based on the recorded operating data, a stranding report can be generated, providing a comprehensive analysis and assessment of the stranding machine's operating status. This allows monitoring of the stranding machine's operating status, timely identification of problems, and implementation of corrective measures.

[0054] It should be understood that before using the above-mentioned neural network model for inference, it is necessary to train the image region feature extractor based on the RCNN model, the local feature context association encoder based on the converter, and the classifier. In other words, the monitoring system 300 for the stranding machine according to the present application further includes a training phase 400 for training the image region feature extractor based on the RCNN model, the local feature context association encoder based on the converter, and the classifier.

[0055] Figure 3 FIG. 1 is a block diagram of the training phase of the monitoring system for the stranding machine according to an embodiment of the present application. Figure 3As shown, the monitoring system 300 of the stranding machine according to the embodiment of the present application includes: a training stage 400, including: a training data acquisition unit 410, for acquiring training data, wherein the training data includes the stranding training speed value, training tension value, training temperature value and training current value at a plurality of predetermined time points within a predetermined time period collected by each monitoring module deployed in the monitored stranding machine; a data arrangement unit 420, for arranging the stranding training speed value, training tension value, training temperature value and training current value at the plurality of predetermined time points into a training stranding speed time series input vector, a training tension time series input vector, a training current time series input vector and a training current time series input vector, respectively, according to the time dimension. The training vector-to-image domain conversion unit 430 is used to convert the training strand speed timing input vector, the training tension timing input vector, the training temperature timing input vector and the training current timing input vector through a vector-to-image converter to obtain a training strand speed timing image, a training tension timing image, a training temperature timing image and a training current timing image; the training image channel aggregation unit 440 is used to convert the training strand speed timing image, the training tension timing image, the training temperature timing image and the training current timing image along the channel dimension. Aggregation to obtain a training parameter multi-channel time series image; a training data time series image local feature extraction unit 450, used to extract features from the training parameter multi-channel time series image through an image region feature extractor based on an RCNN model to obtain a sequence of training parameter multi-channel time series associated semantic feature vectors; a context semantic association feature extraction unit 460, used to pass the sequence of training parameter multi-channel time series associated semantic feature vectors through a local feature context association encoder based on a converter to obtain a training multi-parameter global time series context semantic association feature vector; a classification loss unit 470, used to convert the training multi-parameter global time series context semantic association feature vectors into ... The text semantic association feature vector passes through the classifier to obtain a classification loss function value; the loss function calculation unit 480 is used to calculate the loss function value between the sequence of the training parameter multi-channel temporal association semantic feature vector and the training multi-parameter global temporal context semantic association feature vector; the weighted calculation unit 490 is used to calculate the weighted sum between the loss function value and the classification loss function value as the final loss function value; the training unit 500 is used to train the image area feature extractor based on the RCNN model, the local feature context association encoder based on the converter and the classifier based on the final loss function value.

[0056] Among them, the classification loss unit includes: using the classifier to process the training multi-parameter global temporal context semantic association feature vector to obtain a training classification result; and calculating the cross entropy loss function value between the training classification result and the true value of whether the fault signal is issued as the classification loss function value.

[0057] In particular, in the technical solution of the present application, here, each parameter multi-channel temporal association semantic feature vector in the sequence of parameter multi-channel temporal association semantic feature vectors respectively expresses the local time domain-local time domain temporal association characteristics of the strand speed value, tension value, temperature value and current value in the local time domain of the global time domain determined by vector-image conversion, and each parameter multi-channel temporal association semantic feature vector follows the inter-sample channel distribution of the RCNN model based on multi-sample data. In this way, the sequence of parameter multi-channel temporal association semantic feature vectors is passed through a converter-based local feature context association encoder, and the multi-parameter global temporal context semantic association feature vector obtained can express the mixed time domain spatial scale temporal association feature representation of each parameter value under the inter-sample channel context association. However, considering that under the mixed dimension of sample space channel context association, the multi-parameter global temporal context semantic association feature vector still needs to improve the multi-time domain spatial scale temporal association feature sharing with the sequence of the parameter multi-channel temporal association semantic feature vector, so as to avoid the sparse distribution of key cross-scale temporal features under the time domain-channel space hybrid dimension, the applicant of this application introduces a specific loss function for strengthening temporal feature sharing for the sequence of the parameter multi-channel temporal association semantic feature vector and the multi-parameter global temporal context semantic association feature vector, which is expressed as:

[0058]

[0059] Wherein, V1 is the concatenated feature vector obtained by sequential concatenation of the parameter multi-channel temporal association semantic feature vectors, and V2 is the multi-parameter global temporal context semantic association feature vector, ‖·‖1 and ‖·‖2 are the 1-norm and 2-norm of the feature vector respectively, ε is the threshold hyperparameter, and the feature vectors are all in the form of row vectors, Indicates difference by position. Represents vector multiplication. Specifically, the reinforcement of the shared key temporal features between the sequence of the parameter multi-channel temporal association semantic feature vectors and the multi-parameter global temporal context semantic association feature vectors can be regarded as the distribution information compression of the global feature set, and the distribution sparse control of the key features is performed on the basis of reconstructing the relative shape relationship of the original feature manifold based on the structural representation between the sequence of the parameter multi-channel temporal association semantic feature vectors and the multi-parameter global temporal context semantic association feature vectors. While strengthening the shared key temporal features between the sequence of the parameter multi-channel temporal association semantic feature vectors and the multi-parameter global temporal context semantic association feature vectors, a sparse but meaningful geometric representation of the fused manifold of the multi-parameter global temporal context semantic association feature vector can be obtained to improve the expression effect of the multi-parameter global temporal context semantic association feature vector, thereby improving the accuracy of the classification result obtained by the classifier. In this way, real-time monitoring and automatic control of the operating status of the stranding machine can be achieved, thereby ensuring the quality of the stranding, improving production efficiency, and reducing failures and losses.

[0060] As described above, the monitoring system 300 for a wire stranding machine according to an embodiment of the present application can be implemented in various wireless terminals, such as a server equipped with a monitoring algorithm for the wire stranding machine. In one possible implementation, the monitoring system 300 for a wire stranding machine according to an embodiment of the present application can be integrated into a wireless terminal as a software module and / or a hardware module. For example, the monitoring system 300 for a wire stranding machine 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 monitoring system 300 for a wire stranding machine can also be one of the many hardware modules of the wireless terminal.

[0061] Alternatively, in another example, the monitoring system 300 of the stranding machine and the wireless terminal may be separate devices, and the monitoring system 300 of the stranding machine may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in an agreed data format.

[0062] Furthermore, a monitoring method for a stranding machine is also provided.

[0063] Figure 7 FIG. 1 is a flow chart of a monitoring method for a stranding machine according to an embodiment of the present application. Figure 7As shown, the monitoring method of the stranding machine according to the embodiment of the present application includes the following steps: S1, measuring the speed and tension of the monitored stranding machine, and synchronously controlling the speed ratio of each axis; S2, measuring the temperature of the monitored stranding machine, and adjusting the switch and air volume of the cooling fan; S3, measuring the current of the monitored stranding machine, and judging the working status and load condition of the stranding machine; S4, detecting the fault signal of the monitored stranding machine, and displaying the fault signal; S5, recording the operating data of the monitored stranding machine, and generating a stranding report and statistical analysis.

[0064] In summary, the monitoring method of the stranding machine according to the embodiment of the present application is explained, which collects various data parameters of the stranding machine in real time during operation by deploying various data monitoring modules on the stranding machine, such as a speed monitoring module, a temperature monitoring module, and a current detection module, and uses a fault diagnosis module to perform a time series collaborative analysis of these stranding machine operation data using data processing and analysis algorithms, so as to determine whether the operation status of the stranding machine is normal, and promptly sends and displays a fault signal when an abnormality is detected, so that corresponding measures can be taken to repair it later, avoiding production interruptions and losses. In this way, real-time monitoring and automatic control of the operation status of the stranding machine can be achieved, thereby ensuring the quality of the stranded wire, improving production efficiency, and reducing failures and losses.

[0065] While various embodiments of the present disclosure have been described above, the above descriptions are illustrative, non-exhaustive, and not intended to be limiting of the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is selected to best explain the principles of the embodiments, their practical applications, or improvements to existing technologies, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A monitoring system for a stranding machine, characterized in that: include: Speed ​​monitoring module, used to measure the speed and tension of the monitored stranding machine, and synchronously control the speed ratio of each axis; A temperature monitoring module, used to measure the temperature of the monitored stranding machine and adjust the switch and air volume of the cooling fan; A current monitoring module, used to measure the current of the monitored stranding machine and determine the working state and load condition of the stranding machine; a fault diagnosis module, configured to detect a fault signal of the monitored stranding machine and display the fault signal; A data recording module for recording the operating data of the monitored stranding machine and generating stranding reports and statistical analysis; Wherein, the fault diagnosis module includes: a data acquisition unit, configured to acquire stranding speed values, tension values, temperature values, and current values ​​at a plurality of predetermined time points within a predetermined time period, collected by various monitoring modules deployed on the monitored stranding machine; a data parameter timing arrangement unit, configured to arrange the strand speed values, tension values, temperature values, and current values ​​at the plurality of predetermined time points into a strand speed timing input vector, a tension timing input vector, a temperature timing input vector, and a current timing input vector, respectively, according to a time dimension; a data preprocessing and aggregation unit, configured to perform image domain conversion and parameter multi-channel aggregation processing on the strand speed time series input vector, the tension time series input vector, the temperature time series input vector, and the current time series input vector to obtain a parameter multi-channel time series image; A data time series image local feature extraction unit is used to extract features from the parameter multi-channel time series image using an image region feature extractor based on a deep neural network model to obtain a sequence of parameter multi-channel time series associated semantic feature vectors; A data temporal image local semantic association coding unit is used to perform context association coding on the sequence of the parameter multi-channel temporal association semantic feature vectors to obtain a multi-parameter global temporal context semantic association feature; A fault signal detection unit is used to determine whether to issue the fault signal based on the multi-parameter global temporal context semantic association feature.

2. The monitoring system for a stranding machine according to claim 1, characterized in that: The data preprocessing and aggregation unit includes: a vector-image domain conversion subunit, configured to convert the strand speed timing input vector, the tension timing input vector, the temperature timing input vector, and the current timing input vector into a vector-image converter to obtain a strand speed timing image, a tension timing image, a temperature timing image, and a current timing image, respectively; The image channel aggregation subunit is used to aggregate the strand speed time series image, the tension time series image, the temperature time series image and the current time series image along the channel dimension to obtain the parameter multi-channel time series image.

3. The monitoring system for a stranding machine according to claim 2, characterized in that: The deep neural network model is an RCNN model.

4. The monitoring system for a stranding machine according to claim 3, characterized in that: The data temporal image local semantic association encoding unit is used to: pass the sequence of the parameter multi-channel temporal association semantic feature vectors through a converter-based local feature context association encoder to obtain a multi-parameter global temporal context semantic association feature vector as the multi-parameter global temporal context semantic association feature.

5. The monitoring system for a stranding machine according to claim 4, characterized in that: The fault signal detection unit includes: a stranding machine operating state detection subunit, configured to pass the multi-parameter global temporal context semantic association feature vector through a classifier to obtain a classification result, wherein the classification result is used to indicate whether the operating state of the monitored stranding machine is abnormal; The fault signal issuing judgment subunit is used to determine whether to issue the fault signal based on the classification result.

6. The monitoring system for a stranding machine according to claim 5, characterized in that: It also includes a training module for training the image region feature extractor based on the RCNN model, the converter-based local feature context association encoder and the classifier.

7. The monitoring system for a stranding machine according to claim 6, characterized in that: The training module includes: a training data acquisition unit, configured to acquire training data, the training data comprising stranding training speed values, training tension values, training temperature values, and training current values ​​at a plurality of predetermined time points within a predetermined time period, collected by various monitoring modules deployed on the monitored stranding machine; a data arranging unit, configured to arrange the strand training speed values, training tension values, training temperature values, and training current values ​​at the plurality of predetermined time points into a training strand speed timing input vector, a training tension timing input vector, a training temperature timing input vector, and a training current timing input vector, respectively, according to a time dimension; a training vector-to-image domain conversion unit, configured to convert the training strand speed timing input vector, the training tension timing input vector, the training temperature timing input vector, and the training current timing input vector into a vector-to-image converter to obtain a training strand speed timing image, a training tension timing image, a training temperature timing image, and a training current timing image, respectively; a training image channel aggregation unit, configured to aggregate the training strand speed time series image, the training tension time series image, the training temperature time series image, and the training current time series image along a channel dimension to obtain a training parameter multi-channel time series image; A training data time series image local feature extraction unit is used to extract features from the training parameter multi-channel time series image using an image region feature extractor based on an RCNN model to obtain a sequence of training parameter multi-channel time series associated semantic feature vectors; a contextual semantic association feature extraction unit, configured to pass the sequence of the training parameter multi-channel temporal association semantic feature vectors through a converter-based local feature contextual association encoder to obtain a training multi-parameter global temporal contextual semantic association feature vector; A classification loss unit, configured to pass the training multi-parameter global temporal context semantic association feature vector through a classifier to obtain a classification loss function value; A loss function calculation unit, configured to calculate a loss function value between the sequence of the training parameter multi-channel temporal association semantic feature vectors and the training multi-parameter global temporal context semantic association feature vector; A weighted calculation unit, configured to calculate a weighted sum of the loss function value and the classification loss function value as a final loss function value; A training unit is used to train the RCNN model-based image region feature extractor, the converter-based local feature context association encoder and the classifier based on the final loss function value.

8. The monitoring system for a stranding machine according to claim 7, characterized in that: The classification loss unit includes: Processing the training multi-parameter global temporal context semantic association feature vector using the classifier to obtain a training classification result; and A cross entropy loss function value between the training classification result and the true value of whether the fault signal is issued is calculated as the classification loss function value.

9. A method for monitoring a wire stranding machine using the monitoring system for a wire stranding machine according to claim 1, characterized in that: include: Measure the speed and tension of the monitored stranding machine and synchronously control the speed ratio of each axis; Measuring the temperature of the monitored stranding machine and adjusting the switch and air volume of the cooling fan; Measuring the current of the monitored stranding machine and determining the working state and load condition of the stranding machine; detecting a fault signal of the monitored stranding machine and displaying the fault signal; Record the operating data of the monitored stranding machine, and generate stranding reports and statistical analysis.

Images