Intelligent control system and method for metal scrap treatment

Abstract

The invention relates to the field of intelligent control, and provides an intelligent control system and method for metal waste treatment, and the method comprises the steps: carrying out the correlation analysis of a metal waste crushing monitoring video and metal waste feeding speed time sequence data, so as to know the fusion characteristics of metal waste crushing rotating speed correlation factors; and meanwhile, time sequence analysis is conducted on crushing rotating speed time sequence data of crushing equipment to know metal waste crushing rotating speed time sequence characteristics, metal waste crushing rotating speed associated factor fusion characteristics and the metal waste crushing rotating speed time sequence characteristics are comprehensively considered, and the crushing rotating speed regulation and control direction of the next time point is obtained through analysis and classification. In this way, regulation and control of the rotating speed of the crushing equipment can be optimized, and therefore intelligent control over the metal scrap crushing treatment process is achieved.

Description

Technical Field

[0001] The present application relates to the field of intelligent control, and more specifically, to an intelligent control system and method for metal waste processing. Background Art

[0002] Metal waste treatment is an important environmental protection and resource recycling work. Metal waste treatment not only reduces the exploitation of natural resources and energy consumption, but also reduces the pollution of waste to the environment, which has important economic and environmental significance.

[0003] In the process of metal scrap processing, the speed of the crushing equipment directly affects the crushing efficiency and energy consumption. Too high a speed may cause increased equipment wear and increased energy consumption, while too low a speed will affect the crushing efficiency and extend the processing time. Traditional control methods are often based on experience or fixed set values, which are difficult to flexibly adjust according to real-time conditions, making it difficult to achieve the best balance between efficiency and energy consumption.

[0004] Therefore, an optimized intelligent control solution for metal waste processing is needed. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present application provides an intelligent control system and method for metal waste processing.

[0006] An intelligent control method for metal waste processing, comprising:

[0007] Obtaining the metal scrap feeding speed values, the crushing speed values ​​of the crushing equipment at a plurality of predetermined time points within a predetermined time period, and the metal scrap crushing monitoring video within the predetermined time period;

[0008] Processing the metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video in the predetermined time period to obtain a fusion feature vector of metal scrap crushing speed correlation factors;

[0009] Performing data processing on the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points to obtain a time series characteristic vector of the crushing speed of the scrap metal;

[0010] The metal scrap crushing speed correlation factor fusion feature vector and the metal scrap crushing speed time series feature vector are fused to obtain a metal scrap crushing responsiveness correlation vector;

[0011] Based on the scrap metal crushing responsiveness association vector, the control direction of the scrap metal crushing rotation speed value at the next time point is obtained.

[0012] An intelligent control system for metal waste processing, comprising:

[0013] A metal scrap crushing processing data collection module, used to obtain the metal scrap feeding speed value, the crushing speed value of the crushing equipment and the metal scrap crushing monitoring video of the predetermined time period at multiple predetermined time points within a predetermined time period;

[0014] A metal scrap crushing speed correlation factor fusion feature extraction module is used to process the metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video of the predetermined time period to obtain a metal scrap crushing speed correlation factor fusion feature vector;

[0015] A metal scrap crushing speed time series feature extraction module, used for performing data processing on the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points to obtain a metal scrap crushing speed time series feature vector;

[0016] A metal scrap crushing speed-related feature fusion module, used for fusing the metal scrap crushing speed-related factor fusion feature vector and the metal scrap crushing speed time series feature vector to obtain a metal scrap crushing responsiveness association vector;

[0017] The metal scrap crushing speed control result generation module is used to obtain the control direction of the metal scrap crushing speed value at the next time point based on the metal scrap crushing responsiveness association vector.

[0018] This application has significant technical effects due to the adoption of the above technical solutions:

[0019] The intelligent control system and method for metal scrap processing provided by the present application performs correlation analysis on the monitoring video of metal scrap crushing and the time series data of metal scrap feeding speed to understand the fusion characteristics of the metal scrap crushing speed correlation factors, and at the same time performs time series analysis on the crushing speed time series data of the crushing equipment to understand the time series characteristics of the metal scrap crushing speed, and comprehensively considers the fusion characteristics of the metal scrap crushing speed correlation factors and the time series characteristics of the metal scrap crushing speed, so as to analyze and classify and obtain the crushing speed control direction at the next time point. In this way, the speed control of the crushing equipment can be optimized, thereby realizing the intelligent control of the metal scrap crushing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used 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 accompanying drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1Flow chart of an intelligent control method for metal waste processing according to an embodiment of the present application.

[0022] Figure 2 Schematic diagram of data flow of an intelligent control method for metal waste processing according to an embodiment of the present application.

[0023] Figure 3 In the intelligent control method for metal waste processing according to an embodiment of the present application, data processing is performed on the metal waste feeding speed values ​​at the multiple predetermined time points and the metal waste crushing monitoring video of the predetermined time period to obtain a flow chart of the fusion feature vector of the metal waste crushing rotation speed correlation factors.

[0024] Figure 4 The present invention is a flow chart of performing video content feature encoding on the monitoring video of metal scrap crushing in the predetermined time period to obtain a feature vector of metal scrap crushing state change in the intelligent control method for metal scrap processing according to an embodiment of the present application.

[0025] Figure 5 4 is a system block diagram of an intelligent control system for metal waste processing 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 here.

[0027] Metal waste treatment is an important task related to environmental protection and resource recycling. In today's society, with the accelerated advancement of industrialization, the amount of metal waste is increasing day by day. If it is not properly handled, it will affect the ecological balance and sustainable development of mankind.

[0028] In the actual operation process of metal scrap processing, crushing is one of the key links, and the speed setting of the crushing equipment is directly related to the crushing efficiency and energy consumption. When the speed of the crushing equipment is too high, the mechanical parts inside the equipment will wear out due to excessive friction, which will not only shorten the service life of the equipment and increase maintenance costs, but also significantly increase energy consumption, resulting in unnecessary energy waste. On the contrary, if the speed is too low, the crushing capacity of the equipment cannot be fully utilized, resulting in a significant reduction in crushing efficiency, thereby extending the entire metal scrap processing time and reducing work efficiency. Traditional methods of speed control of crushing equipment mostly rely on the experience of the operator or fixed set values. This method often seems powerless when faced with complex and changeable actual working conditions.

[0029] In recent years, deep learning and neural networks have been widely used in computer vision, natural language processing, text signal processing and other fields. In addition, deep learning and neural networks have also shown a level close to or even beyond that of humans in image classification, object detection, semantic segmentation, text translation and other fields. The development of deep learning and neural networks provides new solutions and solutions for intelligent control of metal waste processing.

[0030] Figure 1 Flow chart of an intelligent control method for metal waste processing according to an embodiment of the present application. Figure 2 Schematic diagram of data flow of the intelligent control method for metal waste processing according to an embodiment of the present application. Figure 1 and Figure 2 As shown, the intelligent control method for metal scrap processing according to the embodiment of the present application includes: S110, obtaining the metal scrap feeding speed values, the crushing rotation speed values ​​of the crushing equipment and the metal scrap crushing monitoring video of the predetermined time period at multiple predetermined time points within a predetermined time period; S120, performing data processing on the metal scrap feeding speed values ​​at the multiple predetermined time points and the metal scrap crushing monitoring video of the predetermined time period to obtain a fusion feature vector of metal scrap crushing rotation speed correlation factors; S130, performing data processing on the crushing rotation speed values ​​of the crushing equipment at the multiple predetermined time points to obtain a metal scrap crushing rotation speed time series feature vector; S140, fusing the metal scrap crushing rotation speed correlation factor fusion feature vector and the metal scrap crushing rotation speed time series feature vector to obtain a metal scrap crushing responsiveness correlation vector; S150, obtaining the control direction of the metal scrap crushing rotation speed value at the next time point based on the metal scrap crushing responsiveness correlation vector.

[0031] In step S110, the metal scrap feeding speed values, the crushing speed values ​​of the crushing equipment, and the metal scrap crushing monitoring video of the predetermined time period at multiple predetermined time points within a predetermined time period are obtained. It should be understood that the metal scrap feeding speed is one of the key factors affecting the crushing efficiency and energy consumption. When the feeding speed is fast, the crushing equipment needs a higher speed to process more scrap; when the feeding speed is slow, the excessively high speed may cause increased equipment wear and increased energy consumption. The metal scrap crushing monitoring video provides real-time status information of the metal scrap during the crushing process. This data contains key information such as the flow of the scrap in the crushing equipment and the crushing state. When the scrap crushing state is not good, the speed needs to be increased to improve the crushing efficiency. The speed of the crushing equipment directly affects its crushing capacity and energy consumption. When the speed is too high, it will cause increased equipment wear and increased energy consumption; and when the speed is too low, it will affect the crushing efficiency. In general, the timing data of the metal scrap feeding speed value, the timing data of the crushing speed value of the crushing equipment, and the metal scrap crushing monitoring video data are the basis for realizing intelligent control of the crushing equipment speed. These three data respectively contain key information about the waste feed rate, equipment rotation speed and crushing status. By analyzing and integrating this information, the rotation speed can be predicted and adjusted to achieve the best balance between crushing efficiency and energy consumption.

[0032] In step S120, data processing is performed on the metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video of the predetermined time period to obtain a fusion feature vector of metal scrap crushing speed correlation factors. Specifically, Figure 3 The present invention is a flow chart of the intelligent control method for metal waste processing according to the embodiment of the present application, which processes the metal waste feeding speed values ​​at the plurality of predetermined time points and the metal waste crushing monitoring video of the predetermined time period to obtain the metal waste crushing speed correlation factor fusion feature vector. Figure 3 As shown, data processing is performed on the metal scrap feeding speed values ​​at the multiple predetermined time points and the metal scrap crushing monitoring video of the predetermined time period to obtain a fusion feature vector of factors related to metal scrap crushing rotation speed, including: S121, video content feature encoding is performed on the metal scrap crushing monitoring video of the predetermined time period to obtain a metal scrap crushing state change feature vector; S122, time series analysis is performed on the metal scrap feeding speed values ​​at the multiple predetermined time points to obtain a metal scrap feeding speed time series feature vector; S123, internal characteristic interdependence optimization based on invisible spatial modulation is performed on the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector to obtain the metal scrap crushing rotation speed related factor fusion feature vector.

[0033] More specifically, Figure 4The present invention is a flow chart of performing video content feature encoding on the monitoring video of metal scrap crushing in the predetermined time period to obtain a feature vector of the change in the state of metal scrap crushing in the intelligent control method for metal scrap processing according to the embodiment of the present application. Figure 4 As shown, video content feature encoding is performed on the metal scrap crushing monitoring video of the predetermined time period to obtain a metal scrap crushing state change feature vector, including: S1211, sampling the metal scrap crushing monitoring video of the predetermined time period to obtain a plurality of metal scrap crushing key frames; S1212, passing the plurality of metal scrap crushing key frames through a metal scrap crushing state encoder to obtain a metal scrap crushing state change feature graph; S1213, performing mean pooling of each feature matrix along the channel dimension on the metal scrap crushing state change feature graph to obtain the metal scrap crushing state change feature vector.

[0034] In step S1211, the monitoring video of the metal scrap crushing in the predetermined time period is sampled to obtain a plurality of metal scrap crushing key frames. Accordingly, considering that there are a large number of video frames in the monitoring video of the metal scrap crushing in the predetermined time period, directly processing and analyzing the entire video will consume a large amount of data, and also considering that the content difference between adjacent video frames in the crushing monitoring video is small, all processing will process too much redundant information. Therefore, in the technical solution of the present application, it is necessary to sample the monitoring video of the metal scrap crushing in the predetermined time period. By sampling, the key frames in the video can be extracted, thereby greatly reducing the amount of data to be processed, which is convenient for subsequent storage and analysis. In particular, in an embodiment of the present application, a feasible implementation method for sampling the monitoring video of the metal scrap crushing in the predetermined time period to obtain a plurality of metal scrap crushing key frames can be: performing key frame sampling processing on the monitoring video of the metal scrap crushing in the predetermined time period at a predetermined sampling frequency, thereby obtaining a plurality of metal scrap crushing key frames.

[0035] In step S1212, the multiple metal scrap crushing key frames are passed through the metal scrap crushing state encoder to obtain a metal scrap crushing state change characteristic diagram. It should be understood that the multiple metal scrap crushing key frames respectively record the morphological changes of the metal scrap during the crushing process, such as the size and shape of the scrap, which can reflect the degree and uniformity of the crushing of the scrap. In order to more accurately reflect the crushing effect of the metal scrap at different speeds, it is necessary to pass the multiple metal scrap crushing key frames through the metal scrap crushing state encoder. It should be noted that the essence of the metal scrap crushing state encoder here is a convolutional neural network model using a time attention mechanism. Those of ordinary skill in the art should know that the convolutional neural network has a powerful feature extraction capability and can automatically extract useful feature information from the key frames, such as the morphological changes of the metal scrap. By using the time attention mechanism, the model's attention to a specific time period in the input data can be strengthened, thereby more accurately capturing the key information in the metal scrap crushing process. This helps the model better understand the changing laws of the crushing process, thereby improving the accuracy of capturing the metal scrap crushing state.

[0036] Specifically, in an embodiment of the present application, a possible implementation method of passing the multiple metal scrap crushing key frames through a metal scrap crushing state encoder to obtain a metal scrap crushing state change characteristic map may be: extracting adjacent first metal scrap crushing key frames and second metal scrap crushing key frames from the multiple metal scrap crushing key frames; passing the first metal scrap crushing key frame and the second metal scrap crushing key frame through the first convolution layer and the second convolution layer of the metal scrap crushing state encoder respectively to obtain a first metal scrap crushing state convolution feature map and a second metal scrap crushing state convolution feature map; calculating the first metal scrap crushing state convolution feature map and the first metal scrap crushing state convolution feature map. The position point multiplication between the two metal scrap crushing state convolution feature maps is performed to obtain a metal scrap crushing state time attention map; the metal scrap crushing state time attention map is input into the Softmax activation function to obtain the metal scrap crushing state time attention feature map; the second metal scrap crushing key frame passes through the third convolution layer of the metal scrap crushing state encoder to obtain a third metal scrap crushing state convolution feature map; the position point multiplication between the third metal scrap crushing state convolution feature map and the metal scrap crushing state time attention feature map is calculated to obtain the metal scrap crushing state time attention feature map corresponding to the second metal scrap crushing key frame.

[0037] In step S1213, the mean pooling of each feature matrix along the channel dimension is performed on the feature map of the metal scrap crushing state change to obtain the feature vector of the metal scrap crushing state change. It should be understood that the feature map of the metal scrap crushing state change usually contains a large amount of feature information, and this feature information presents a complex distribution in space. However, in the subsequent processing process, all feature information is often not needed, but the most representative features need to be extracted therefrom. Therefore, it is necessary to perform a pooling operation on the feature map of the metal scrap crushing state change, which can effectively reduce the dimension of the feature map while retaining the key feature information.

[0038] In step S122, the metal scrap feed speed values ​​at the plurality of predetermined time points are subjected to a time series analysis to obtain a metal scrap feed speed time series feature vector. Specifically, in an embodiment of the present application, the metal scrap feed speed values ​​at the plurality of predetermined time points are subjected to a time series analysis to obtain a metal scrap feed speed time series feature vector, including: arranging the metal scrap feed speed values ​​at the plurality of predetermined time points into a metal scrap feed speed time series input vector according to the time dimension, and then passing the metal scrap feed speed time series encoder to obtain the metal scrap feed speed time series feature vector. It should be understood that in the metal scrap processing process, the feed speed is an important variable, which directly affects the crushing effect and production efficiency of the crusher. By arranging the feed speed values ​​at different time points into a metal scrap feed speed time series input vector, it can help understand the changing trend of the feed speed, which provides a basis for subsequent time series feature extraction. Accordingly, considering that the time series feature information in the time series input vector of the metal scrap feed speed can reflect the changing law and potential pattern of the feed speed, which is closely related to the subsequent crushing speed control, the arranged metal scrap feed speed time series input vector is passed through the metal scrap feed speed time series encoder. Here, the metal scrap feed speed time series encoder is essentially a convolutional neural network model using a one-dimensional convolution kernel. The one-dimensional convolutional neural network has significant advantages in processing time series data. It can capture local patterns and features in the sequence, such as trends, periodicity, mutations, etc. Through the one-dimensional convolution operation, the model can extract the key information in the time series and convert it into a higher-level feature representation. Specifically, the metal scrap feed speed time series encoder can capture local patterns such as sudden increase or decrease and periodic fluctuations in the feed speed input sequence, and can also learn the correlation between different time points in the time series, which helps to understand how the feed speed changes over time and predict the future feed speed change trend. These key feature information are all included in the encoded metal scrap feed speed time series feature vector.

[0039] In step S123, the metal scrap crushing state change feature vector and the metal scrap feed speed time series feature vector are optimized based on the internal characteristics interdependence of invisible spatial modulation to obtain the metal scrap crushing speed correlation factor fusion feature vector. It should be understood that the metal scrap crushing process is a complex process, in which the metal scrap crushing state and feed speed are both important factors affecting the crushing speed. In order to comprehensively consider the combined influence of the crushing state and feed speed on the crushing speed, thereby improving the comprehensiveness and accuracy of speed control, it is necessary to fuse the metal scrap crushing state change feature vector and the metal scrap feed speed time series feature vector in this application.

[0040] In particular, considering that when the feature vector of the change in the state of the metal scrap crushing is fused with the time series feature vector of the metal scrap feeding speed, the two information sources are interrelated in the physical process, but may have different statistical characteristics and dynamic change patterns in data expression. For example, the feature vector of the change in the state of the metal scrap crushing may more reflect the dynamic change and energy distribution of the state of the metal scrap during the crushing process, while the time series feature vector of the metal scrap feeding speed reflects the time series change of the metal scrap feeding process. Such different data sources and processing methods lead to the heterogeneity of the feature vector in the feature space, which will lead to the feature manifold and feature space mismatch caused by the heterogeneity when the feature fusion of the feature vector of the change in the state of the metal scrap crushing and the time series feature vector of the metal scrap feeding speed is performed, which will cause the fused feature vector to fail to have the smoothness of the feature manifold, thereby affecting the certainty of its feature expression. Therefore, in the technical solution of the present application, the internal characteristic interdependence optimization based on invisible space modulation is performed on the feature vector of the change in the state of the metal scrap crushing and the time series feature vector of the metal scrap feeding speed to obtain the fusion feature vector of the correlation factor of the metal scrap crushing speed.

[0041] Specifically, in an embodiment of the present application, the internal characteristic interdependence optimization based on invisible spatial modulation is performed on the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector to obtain the metal scrap crushing speed correlation factor fusion feature vector, including: constructing a metal scrap crushing speed correlation factor distance topological matrix between the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector; constructing a metal scrap crushing speed correlation factor eigenvalue granularity correlation matrix between the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector; based on the metal scrap crushing speed correlation factor distance topological matrix, the metal scrap The eigenvalue granularity association matrix of the metal scrap crushing speed association factors is topologically modulated to obtain the topological modulation association matrix of the metal scrap crushing speed association factors; the topological modulation association matrix of the metal scrap crushing speed association factors is used as the perspective co-projection space, and the metal scrap crushing state change characteristic vector and the metal scrap feeding speed time series characteristic vector are projected into the perspective co-projection space to obtain the perspective-modulated metal scrap crushing state change characteristic vector and the perspective-modulated metal scrap feeding speed time series characteristic vector; the perspective-modulated metal scrap crushing state change characteristic vector and the perspective-modulated metal scrap feeding speed time series characteristic vector are fused to obtain the metal scrap crushing speed association factor fusion characteristic vector.

[0042] In an embodiment of the present application, another possible implementation method of performing internal characteristic interdependence optimization based on invisible spatial modulation on the scrap metal crushing state change feature vector and the scrap metal feeding speed time series feature vector to obtain a scrap metal crushing speed correlation factor fusion feature vector may be: the scrap metal crushing state change feature vector and the scrap metal feeding speed time series feature vector are processed by the following formula to obtain a scrap metal crushing speed correlation factor fusion feature vector; wherein the formula is:

[0043] M t =D(V1, V2)

[0044]

[0045]

[0046]

[0047]

[0048] V c =α·V1′+β·V2′

[0049] Wherein, V1 represents the characteristic vector of the state change of the metal scrap crushing, V2 represents the time series characteristic vector of the metal scrap feeding speed, and D(V1, V2) is the distance topological matrix of the metal scrap crushing speed correlation factor between the metal scrap crushing state change characteristic vector and the metal scrap feeding speed time series characteristic vector, that is, d i，j =d(v 1i , v 2j ), d i，j is the eigenvalue of the (i, j)th position in the distance topological matrix of the metal scrap crushing speed correlation factor, d(v 1i , v 2j ) represents the distance between the eigenvalue of the ith position of the metal scrap crushing state change eigenvalue and the eigenvalue of the jth position of the metal scrap feeding speed time series eigenvalue, and V1 and V2 are both column vectors, M t represents the distance topological matrix of the metal scrap crushing speed correlation factor, T represents the transpose of the vector, Represents matrix multiplication, M v represents the eigenvalue granularity correlation matrix of the metal scrap crushing speed correlation factor, M c represents the topological modulation correlation matrix of the metal scrap crushing speed correlation factors, exp represents the natural exponential function, V1′ represents the perspective modulation metal scrap crushing state change feature vector, V2′ represents the perspective modulation metal scrap feeding speed time series feature vector, α and β represent weighted hyperparameters, V c It represents the fusion feature vector of the metal scrap crushing rotation speed correlation factors.

[0050] That is, in the technical solution of the present application, the perspective projection matching based on the feature topological association is performed on the feature vector of the state change of the metal scrap crushing and the feature vector of the time series of the metal scrap feeding speed, which first constructs the distance topological matrix of the metal scrap crushing rotation speed association factor between the feature vector of the state change of the metal scrap crushing and the feature vector of the time series of the metal scrap feeding speed. This step utilizes the idea of ​​graph theory, takes the feature values ​​of each position in the feature vector of the state change of the metal scrap crushing and the feature vector of the time series of the metal scrap feeding speed as nodes, and uses the distance metric function to determine the low-dimensional embedding expression of the edges between nodes, so as to construct a feature expression that can quantify the topological association relationship between the feature vector of the state change of the metal scrap crushing and the feature vector of the time series of the metal scrap feeding speed. This step is crucial to understanding the spatial distribution of data points and their mutual connection, and also provides basic structural information for subsequent steps.

[0051] Next, a metal scrap crushing speed correlation factor characteristic value granularity correlation matrix between the metal scrap crushing state change characteristic vector and the metal scrap feeding speed time series characteristic vector is constructed. This step focuses on exploring the correlation between the internal attributes of the characteristic vector. In a specific example, the product between the metal scrap crushing state change characteristic vector and the transposed vector of the metal scrap feeding speed time series characteristic vector is calculated to obtain the metal scrap crushing speed correlation factor characteristic value granularity correlation matrix. Furthermore, the metal scrap crushing speed correlation factor characteristic value granularity correlation matrix is ​​topologically correlated modulated based on the metal scrap crushing speed correlation factor distance topological matrix to obtain the metal scrap crushing speed correlation factor topological modulation correlation matrix. That is, the characteristic value granularity correlation information of the metal scrap crushing state change characteristic vector and the metal scrap feeding speed time series characteristic vector is driven to wander on the distance topological map to construct a low-dimensional modulated perspective co-projection space domain for mapping the metal scrap crushing state change characteristic vector and the metal scrap feeding speed time series characteristic vector to a continuous high-dimensional regression space attribute.

[0052] Next, the topological modulation correlation matrix of the metal scrap crushing speed correlation factor is used as the perspective co-projection space, and the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector are projected into the perspective co-projection space to obtain the perspective modulation metal scrap crushing state change feature vector and the perspective modulation metal scrap feeding speed time series feature vector. That is, the key to this technical solution is to find an implicit third space with modulation capability, in which as much information of the original high-dimensional data as possible is retained, so that the converted new feature vector not only reduces the dimension, but also maintains the relative position relationship between the original features.

[0053] Finally, the perspective-modulated scrap metal crushing state change feature vector and the perspective-modulated scrap metal feed speed time series feature vector are fused to obtain the scrap metal crushing rotation speed correlation factor fusion feature vector. Fusion can be achieved in a variety of forms, ranging from simple arithmetic averages to complex ensemble learning algorithms, depending on the application scenario and the goals pursued. The final scrap metal crushing rotation speed correlation factor fusion feature vector combines the main features of the two sets of input data and provides a more comprehensive data representation, which is not only conducive to improving the performance of machine learning models, but also promotes the ability of cross-modal or multi-source data analysis, thereby significantly enhancing the understanding and utilization efficiency of complex data structures.

[0054] In step S130, the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points are processed to obtain a metal scrap crushing speed time series feature vector. Specifically, in an embodiment of the present application, the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points are processed to obtain a metal scrap crushing speed time series feature vector, including: arranging the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points into a metal scrap crushing speed time series input vector according to the time dimension and then passing through a crushing speed time series encoder to obtain the metal scrap crushing speed time series feature vector. It should be understood that the crushing speed of the crushing equipment is a dynamic parameter that changes with time. In order to capture the subtle changes and potential trends in the crushing speed during the metal scrap process, and then to deeply understand the dynamic characteristics of the speed during the crushing process, it is necessary to extract time series features from the crushing speed time series data. Specifically, first, the crushing speed data at multiple predetermined time points are arranged into a metal scrap crushing speed time series input vector according to the time dimension to facilitate the subsequent sequence data feature extraction, and then the crushing speed time series encoder is used to process the arranged metal scrap crushing speed time series input vector to obtain a metal scrap crushing speed time series feature vector containing the periodic change trend information of the crushing speed. It should be noted that the crushing speed time series encoder here is a convolutional neural network model using a one-dimensional convolution kernel.

[0055] In step S140, the fusion feature vector of the metal scrap crushing speed correlation factor and the metal scrap crushing speed time series feature vector are fused to obtain the metal scrap crushing responsiveness correlation vector. It should be understood that the fusion feature vector of the metal scrap crushing speed correlation factor contains the characteristics of various key factors affecting the crushing speed, while the metal scrap crushing speed time series feature vector reflects the dynamic characteristics of the crushing speed changing over time. In order to obtain richer feature information and more accurately describe the material state and trend of the crushing process, it is necessary to fuse the fusion feature vector of the metal scrap crushing speed correlation factor and the metal scrap crushing speed time series feature vector in the technical solution of the present application to obtain the metal scrap crushing responsiveness correlation vector. In particular, in the embodiment of the present application, one possible implementation method for fusing the metal scrap crushing speed correlation factor fusion feature vector and the metal scrap crushing speed time series feature vector can be: multiplying the metal scrap crushing speed correlation factor fusion feature vector and the metal scrap crushing speed time series feature vector by position point to obtain the metal scrap crushing responsiveness correlation vector.

[0056] In step S150, based on the scrap metal crushing responsiveness association vector, the control direction of the scrap metal crushing speed value at the next time point is obtained. Specifically, in the embodiment of the present application, based on the scrap metal crushing responsiveness association vector, the control direction of the scrap metal crushing speed value at the next time point is obtained, including: passing the scrap metal crushing responsiveness association vector through a crushing speed classifier to obtain a classification result, and the classification result is used to indicate whether the crushing speed value at the next time point should increase, decrease, or remain unchanged. It should be understood that the scrap metal crushing responsiveness association vector is a complex feature representation, and it is not intuitive enough to be directly used for speed control. In order to map this vector to a detailed and easy-to-understand classification result, it is necessary to pass the scrap metal crushing responsiveness association vector through a crushing speed classifier. The crushing speed classifier is a machine learning model, which can output a classification result indicating that the crushing speed value at the next time point should increase, decrease, or remain unchanged according to the input scrap metal crushing responsiveness association vector through a large amount of data training. The classification result can be directly used to guide the speed control of the crushing equipment, thereby optimizing the crushing efficiency and energy consumption in the scrap metal crushing process.

[0057] In summary, the intelligent control method for metal scrap processing based on the embodiment of the present application is explained, which performs correlation analysis on the monitoring video of metal scrap crushing and the time series data of metal scrap feeding speed to understand the fusion characteristics of metal scrap crushing speed correlation factors, and simultaneously performs time series analysis on the crushing speed time series data of the crushing equipment to understand the time series characteristics of the metal scrap crushing speed, and comprehensively considers the fusion characteristics of the metal scrap crushing speed correlation factors and the time series characteristics of the metal scrap crushing speed, so as to analyze and classify and obtain the crushing speed control direction at the next time point. In this way, the speed control of the crushing equipment can be optimized, thereby realizing the intelligent control of the metal scrap crushing process.

[0058] Figure 5 : is a system block diagram of an intelligent control system for metal waste processing according to an embodiment of the present application. Figure 5As shown, the intelligent control system 100 for metal waste processing according to the embodiment of the present application includes: a metal waste crushing processing data collection module 110, which is used to obtain the metal waste feeding speed values, the crushing speed values ​​of the crushing equipment and the metal waste crushing monitoring video of the predetermined time period at multiple predetermined time points within a predetermined time period; a metal waste crushing speed correlation factor fusion feature extraction module 120, which is used to perform data processing on the metal waste feeding speed values ​​at the multiple predetermined time points and the metal waste crushing monitoring video of the predetermined time period to obtain the metal waste crushing speed correlation factor fusion feature vector; The metal scrap crushing speed time series feature extraction module 130 is used to perform data processing on the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points to obtain the metal scrap crushing speed time series feature vector; the metal scrap crushing speed related feature fusion module 140 is used to fuse the metal scrap crushing speed related factor fusion feature vector and the metal scrap crushing speed time series feature vector to obtain the metal scrap crushing responsiveness associated vector; the metal scrap crushing speed control result generation module 150 is used to obtain the control direction of the metal scrap crushing speed value at the next time point based on the metal scrap crushing responsiveness associated vector.

[0059] Here, those skilled in the art will appreciate that the specific functions and operations of the various units and modules in the intelligent control system 100 for metal waste processing have been described above with reference to Figures 1 to 4 The invention has been described in detail in the description of the intelligent control method for metal waste processing, and therefore, its repeated description will be omitted.

[0060] In summary, the intelligent control system 100 for metal waste processing based on the embodiment of the present application is explained, which performs correlation analysis on the monitoring video of metal waste crushing and the time series data of metal waste feeding speed to understand the fusion characteristics of metal waste crushing speed correlation factors, and simultaneously performs time series analysis on the crushing speed time series data of the crushing equipment to understand the time series characteristics of the metal waste crushing speed, comprehensively considers the fusion characteristics of the metal waste crushing speed correlation factors and the time series characteristics of the metal waste crushing speed, and analyzes and classifies to obtain the crushing speed control direction at the next time point. In this way, the speed control of the crushing equipment can be optimized, thereby realizing the intelligent control of the metal waste crushing process.

[0061] As described above, the intelligent control system 100 for metal waste treatment according to the embodiment of the present application can be implemented in various wireless terminals, such as a server for intelligent control of metal waste treatment. In one example, the intelligent control system 100 for metal waste treatment 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 intelligent control system 100 for metal waste treatment 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 intelligent control system 100 for metal waste treatment can also be one of the many hardware modules of the wireless terminal.

[0062] Alternatively, in another example, the intelligent control system 100 for metal waste processing and the wireless terminal may also be separate devices, and the intelligent control system 100 for metal waste processing 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.

Claims

1. An intelligent control method for metal waste processing, characterized in that: include: Obtaining the metal scrap feeding speed values, the crushing speed values ​​of the crushing equipment at a plurality of predetermined time points within a predetermined time period, and the metal scrap crushing monitoring video within the predetermined time period; Processing the metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video in the predetermined time period to obtain a fusion feature vector of metal scrap crushing speed correlation factors; Performing data processing on the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points to obtain a time series characteristic vector of the crushing speed of the scrap metal; The metal scrap crushing speed correlation factor fusion feature vector and the metal scrap crushing speed time series feature vector are fused to obtain a metal scrap crushing responsiveness correlation vector; Based on the scrap metal crushing responsiveness association vector, the control direction of the scrap metal crushing rotation speed value at the next time point is obtained.

2. The intelligent control method for metal waste processing according to claim 1, characterized in that: The metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video in the predetermined time period are processed to obtain a fusion feature vector of metal scrap crushing speed correlation factors, including: Performing video content feature encoding on the monitoring video of metal scrap crushing in the predetermined time period to obtain a feature vector of a change in a state of metal scrap crushing; Performing time series analysis on the metal scrap feeding speed values ​​at the plurality of predetermined time points to obtain a metal scrap feeding speed time series feature vector; The metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector are optimized based on the internal characteristics interdependence of invisible spatial modulation to obtain the metal scrap crushing speed correlation factor fusion feature vector.

3. The intelligent control method for metal waste processing according to claim 2, characterized in that: The video content feature encoding is performed on the monitoring video of the metal scrap crushing in the predetermined time period to obtain a feature vector of the change of the metal scrap crushing state, including: Sampling the metal scrap crushing monitoring video of the predetermined time period to obtain a plurality of metal scrap crushing key frames; Passing the plurality of metal scrap crushing key frames through a metal scrap crushing state encoder to obtain a metal scrap crushing state change characteristic diagram; The metal scrap crushing state change characteristic graph is subjected to mean pooling of each characteristic matrix along the channel dimension to obtain the metal scrap crushing state change characteristic vector.

4. The intelligent control method for metal waste processing according to claim 3 is characterized in that: The metal scrap feeding speed values ​​at the plurality of predetermined time points are subjected to a time series analysis to obtain a metal scrap feeding speed time series characteristic vector, comprising: arranging the metal scrap feeding speed values ​​at the plurality of predetermined time points into a metal scrap feeding speed time series input vector according to the time dimension and then passing the vector through a metal scrap feeding speed time series encoder to obtain the metal scrap feeding speed time series characteristic vector.

5. The intelligent control method for metal waste processing according to claim 4, characterized in that: The metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector are subjected to internal characteristic interdependence optimization based on invisible spatial modulation to obtain the metal scrap crushing speed correlation factor fusion feature vector, including: Constructing a distance topological matrix of metal scrap crushing speed correlation factors between the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector; Constructing a metal scrap crushing speed correlation factor eigenvalue granularity correlation matrix between the metal scrap crushing state change eigenvector and the metal scrap feeding speed time series eigenvector; Based on the distance topological matrix of the metal scrap crushing speed correlation factor, the characteristic value granularity correlation matrix of the metal scrap crushing speed correlation factor is topologically modulated to obtain a metal scrap crushing speed correlation factor topological modulation correlation matrix; Taking the topological modulation correlation matrix of the metal scrap crushing speed correlation factors as the perspective co-projection space, projecting the metal scrap crushing state change feature vector and the metal scrap feeding speed time series feature vector into the perspective co-projection space to obtain the perspective-modulated metal scrap crushing state change feature vector and the perspective-modulated metal scrap feeding speed time series feature vector; The view angle modulated scrap metal crushing state change feature vector and the view angle modulated scrap metal feeding speed time series feature vector are fused to obtain the scrap metal crushing rotation speed associated factor fusion feature vector.

6. The intelligent control method for metal waste processing according to claim 5, characterized in that: The crushing speed values ​​of the crushing equipment at the multiple predetermined time points are processed to obtain a metal scrap crushing speed time series feature vector, including: arranging the crushing speed values ​​of the crushing equipment at the multiple predetermined time points into a metal scrap crushing speed time series input vector according to the time dimension and then passing the vector through a crushing speed time series encoder to obtain the metal scrap crushing speed time series feature vector.

7. The intelligent control method for metal waste processing according to claim 6, characterized in that: The metal scrap feeding speed timing encoder and the crushing rotation speed timing encoder are convolutional neural network models using one-dimensional convolution kernels, and the metal scrap crushing state encoder is a convolutional neural network model using a temporal attention mechanism.

8. The intelligent control method for metal waste processing according to claim 7, characterized in that: Based on the metal scrap crushing responsiveness association vector, the control direction of the metal scrap crushing rotational speed value at the next time point is obtained, including: passing the metal scrap crushing responsiveness association vector through a crushing rotational speed classifier to obtain a classification result, and the classification result is used to indicate that the crushing rotational speed value at the next time point should increase, decrease or remain unchanged.

9. An intelligent control system for metal waste processing, characterized in that: include: A metal scrap crushing processing data collection module, used to obtain the metal scrap feeding speed value, the crushing speed value of the crushing equipment and the metal scrap crushing monitoring video of the predetermined time period at multiple predetermined time points within a predetermined time period; A metal scrap crushing speed correlation factor fusion feature extraction module is used to process the metal scrap feeding speed values ​​at the plurality of predetermined time points and the metal scrap crushing monitoring video of the predetermined time period to obtain a metal scrap crushing speed correlation factor fusion feature vector; A metal scrap crushing speed time series feature extraction module, used for performing data processing on the crushing speed values ​​of the crushing equipment at the plurality of predetermined time points to obtain a metal scrap crushing speed time series feature vector; A metal scrap crushing speed-related feature fusion module, used for fusing the metal scrap crushing speed-related factor fusion feature vector and the metal scrap crushing speed time series feature vector to obtain a metal scrap crushing responsiveness association vector; The metal scrap crushing speed control result generation module is used to obtain the control direction of the metal scrap crushing speed value at the next time point based on the metal scrap crushing responsiveness association vector.

10. The intelligent control system for metal waste processing according to claim 9, characterized in that: The scrap metal crushing speed control result generation module is used to: pass the scrap metal crushing responsiveness association vector through a crushing speed classifier to obtain a classification result, and the classification result is used to indicate whether the crushing speed value at the next time point should increase, decrease or remain unchanged.

Images