Coal flow automatic control system and method for bucket wheel machine

Abstract

According to the automatic coal flow control system and method for the bucket wheel machine, a belt weigher and an encoder which are arranged on a cantilever belt are used as data acquisition equipment, and a deep learning algorithm is used for processing and analyzing a coal flow measurement value acquired by the belt weigher and a cantilever belt pitching angle value acquired by the encoder; the actual measurement value of the coal flow is corrected according to the cantilever belt pitching angle time sequence characteristic information, so that the coal flow measurement value of the belt weigher returns to the normal level, intelligent early warning prompt of belt overload is achieved based on the correction, and equipment damage and safety accidents caused by belt overload are avoided.

Description

Technical Field

[0001] The present application relates to the field of intelligent control technology, and more specifically, to an automatic coal flow control system and method for a bucket wheel excavator. Background Art

[0002] Bucket wheel excavator is a large excavation equipment used for coal mining. Its working principle is to transport coal from the bucket wheel of the bucket wheel excavator to the belt conveyor through a cantilever belt, and then the belt conveyor transports the coal to the designated location. The operating efficiency and safety of the bucket wheel excavator are closely related to the control of coal flow. Excessive coal flow will cause belt overload, resulting in equipment damage and safety accidents. Too small coal flow will reduce production efficiency and resource utilization. Realizing automatic control of coal flow is of great significance to preventing belt overload, and is also an important way to improve the intelligence level of bucket wheel excavators.

[0003] At present, there are some technical solutions for automatic control of coal flow in bucket wheel machines at home and abroad, but they all have certain limitations. For example, some solutions use electronic belt scales on cantilever belts to measure coal flow, but because the pitch angle of the cantilever belt changes with the working state of the bucket wheel machine, the measurement accuracy of the belt scale is poor. Generally speaking, the change of the pitch angle of the cantilever belt will affect the distribution of coal on the belt, resulting in the displacement of the center of gravity of the coal, thus affecting the measurement results of the belt scale.

[0004] Therefore, an optimized solution is expected. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a coal flow automatic control system and method for a bucket wheel excavator, which uses a belt scale and an encoder deployed on a cantilever belt as data acquisition equipment, and uses a deep learning algorithm to process and analyze the coal flow measurement value collected by the belt scale and the cantilever belt pitch angle value collected by the encoder, and corrects the actual measurement value of the coal flow with the cantilever belt pitch angle time series feature information, so that the coal flow measurement value of the belt scale can return to a normal level, and based on this, realizes intelligent early warning prompts for belt overload, so as to avoid equipment damage and safety accidents caused by belt overload.

[0006] In a first aspect, a method for automatically controlling coal flow for a bucket wheel machine is provided, comprising:

[0007] Obtaining coal flow measurement values ​​collected by the belt scale at a plurality of predetermined time points within a predetermined time period;

[0008] Acquire the cantilever belt pitch angle values ​​at a plurality of predetermined time points within the predetermined time period collected by the encoder;

[0009] Performing data preprocessing on the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points to obtain a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector;

[0010] Performing a time series joint analysis on the coal flow measurement value time series input vector and the cantilever belt pitch angle time series input vector to obtain a corrected coal flow measurement value time series feature vector; and

[0011] Based on the time series characteristic vector of the corrected coal flow measurement value, it is confirmed whether a belt overload warning prompt is generated.

[0012] In a second aspect, a coal flow automatic control system for a bucket wheel machine is provided, comprising:

[0013] A coal flow measurement value acquisition module, used to acquire coal flow measurement values ​​collected by a belt scale at a plurality of predetermined time points within a predetermined time period;

[0014] A pitch angle value acquisition module, used to acquire the cantilever belt pitch angle values ​​at a plurality of predetermined time points within the predetermined time period collected by the encoder;

[0015] A data preprocessing module, used for performing data preprocessing on the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points to obtain a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector;

[0016] A time series joint analysis module, used for performing a time series joint analysis on the coal flow measurement value time series input vector and the cantilever belt pitch angle time series input vector to obtain a corrected coal flow measurement value time series feature vector; and

[0017] The belt overload warning prompt determination module is used to confirm whether to generate a belt overload warning prompt based on the time series feature vector of the corrected coal flow measurement value. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 It is a schematic diagram of the technology adopted for the intelligentization of the bucket wheel stacker and reclaimer according to an embodiment of the present application.

[0020] Figure 2 Schematic diagram of a control system network according to an embodiment of the present application.

[0021] Figure 3 It is a flow chart of a method for automatic control of coal flow for a bucket wheel excavator according to an embodiment of the present application.

[0022] Figure 4 Schematic diagram of the architecture of a method for automatic control of coal flow for a bucket wheel excavator according to an embodiment of the present application.

[0023] Figure 5 It is a block diagram of an automatic coal flow control system for a bucket wheel machine according to an embodiment of the present application.

[0024] Figure 6 It is a scenario schematic diagram of the method for automatic control of coal flow for a bucket wheel excavator according to an embodiment of the present application. DETAILED DESCRIPTION

[0025] The following will describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0026] Unless otherwise specified, all technical and scientific terms used in the embodiments of the present application have the same meaning as those commonly understood by those skilled in the art of the present application. The terms used in this application are only for the purpose of describing specific embodiments and are not intended to limit the scope of this application.

[0027] In the embodiments of the present application, it should be noted that, unless otherwise specified and limited, the term "connection" should be understood in a broad sense. For example, it can be an electrical connection or a connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0028] It should be noted that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects, and do not represent a specific order for the objects. It is understandable that the specific order or sequence of "first\second\third" can be interchanged where permitted. It should be understood that the objects distinguished by "first\second\third" can be interchanged where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.

[0029] With the advent of the Industrial 4.0 era, the power industry is undergoing a major transformation. Technologies such as cloud computing, big data, artificial intelligence, and 3D virtual reality have ushered in a new era. Intelligence has become one of the important contents of industry innovation-driven development. Building smart power stations is the development direction of future power station construction.

[0030] Large-scale stacker-reclaimers in thermal power plants are mainly divided into cantilever bucket wheel stackers (including single-function cantilever bucket wheel stackers and cantilever bucket wheel reclaimers), portal bucket wheel stackers, circular stackers, etc. Among them, circular stackers have realized remote operation a few years ago, making them unmanned on site; cantilever bucket wheel stackers are a kind of efficient and continuous loading and unloading machinery, which is most widely used in large thermal power plants. At present, domestic thermal power plants basically use manual on-site operation, with harsh working links, long-term continuous operation, high labor intensity, and working efficiency and safety completely depend on the driver's proficiency and responsibility.

[0031] For coal-fired power plants, blending and burning coal is an effective way to improve economic efficiency. Different types of coal in thermal power plants are piled in the coal yard, and some are stacked in layers. When the boiler draws up a coal stacking and blending plan based on the load and existing coal storage conditions, it guides the bucket wheel machine to perform correct coal stacking and coal taking operations. However, since the bucket wheel machine uses on-site manual operation, especially for the complex coal storage structure of the coal yard, there is a possibility of taking the wrong type of coal, causing the blending plan to fail, affecting economic efficiency and environmental emissions, and seriously causing coking and safety of the boiler. The use of intelligent and automatic operation of the bucket wheel machine can avoid the above situation.

[0032] Compared with gantry bucket wheel stacker and reclaimer and circular stacker and reclaimer, cantilever bucket wheel stacker and reclaimer has the greatest difficulty in intelligent control due to its mechanical structure. With the development of relevant measurement and control technologies, the integration of multiple technologies, such as Figure 1 As shown, it becomes possible to realize unmanned operation of the cantilever bucket wheel stacker and reclaimer.

[0033] To realize the intelligentization of bucket wheel machine, it is necessary to carry out appropriate transformation of its software / hardware. The main hardware transformation contents are as follows: adding precise positioning equipment for bucket wheel machine; laser 3D scanning and anti-collision detection equipment; frequency conversion transformation on bucket wheel machine; adding module I / O channels to PLC on machine; fiber optic communication transformation; server and operator station; video monitoring equipment on bucket wheel machine; fiber optic communication; the network of control system after transformation is as follows: Figure 2 shown.

[0034] Before the stacker and reclaimer is unmanned, the operator manually observes the size of the coal flow and controls it. After the unmanned operation is realized, the human eyes are no longer needed, and the "eyes" of technology must keep up. After the stacker and reclaimer is intelligently unmanned, the automatic control technology of coal flow is an important technology for the intelligent operation of bucket wheel excavators. Its purpose is to prevent belt overload, and sometimes it is required by the coal blending ratio.

[0035] Some bucket wheel excavators are equipped with electronic belt scales on their cantilever belts. However, due to the constant change in the pitch angle of the cantilever belt and the lack of correction methods, the belt scale has extremely poor accuracy during practical use and cannot be used to measure the coal flow control of bucket wheel excavators. In order to achieve a relatively accurate and easy-to-maintain coal flow measurement method, a laser scanner can be installed above the cantilever belt, and the volume can be calculated by scanning the surface shape of the coal flow, and then the density can be estimated and converted into flow. The advantages of this method are less error drift, relative stability, and less frequent correction. The disadvantages are that the density of coal requires an empirical value, and secondly, serious running of the cantilever belt will affect the calculation accuracy of the coal flow cross-sectional area, and an effective correction device needs to be installed. In actual use, the drive motor current of the wheel bucket is introduced to participate in the coal flow control, and suitable control parameters need to be found during the actual debugging process.

[0036] In one embodiment of the present application, Figure 3 It is a flow chart of a method for automatic control of coal flow for a bucket wheel excavator according to an embodiment of the present application. Figure 4 FIG. 1 is a schematic diagram of the structure of the method for automatically controlling the coal flow rate of a bucket wheel excavator according to an embodiment of the present application. Figure 3 and Figure 4 As shown, the automatic control method of coal flow for bucket wheel excavator comprises: 110, obtaining coal flow measurement values ​​at multiple predetermined time points within a predetermined time period collected by a belt scale; 120, obtaining cantilever belt pitch angle values ​​at multiple predetermined time points within the predetermined time period collected by an encoder; 130, performing data preprocessing on the coal flow measurement values ​​at the multiple predetermined time points and the cantilever belt pitch angle values ​​at the multiple predetermined time points to obtain a coal flow measurement value timing input vector and a cantilever belt pitch angle timing input vector; 140, performing timing joint analysis on the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector to obtain a corrected coal flow measurement value timing feature vector; and, 150, based on the corrected coal flow measurement value timing feature vector, confirming whether a belt overload warning prompt is generated.

[0037] In response to the above technical problems, the technical concept of the present application is to use the belt scale and encoder deployed on the cantilever belt as data acquisition equipment, and use a deep learning algorithm to process and analyze the coal flow measurement value collected by the belt scale and the cantilever belt pitch angle value collected by the encoder, and use the cantilever belt pitch angle timing characteristic information to correct the actual measurement value of the coal flow, so that the coal flow measurement value of the belt scale can return to a normal level, and based on this, realize intelligent early warning prompts for belt overload, so as to avoid equipment damage and safety accidents caused by belt overload.

[0038] Based on this, in the technical solution of the present application, firstly, the coal flow measurement values ​​at multiple predetermined time points within a predetermined time period collected by the belt scale are obtained; and the cantilever belt pitch angle values ​​at multiple predetermined time points within the predetermined time period collected by the encoder are obtained. That is, the coal flow measurement values ​​at multiple predetermined time points are collected, and the transportation of coal is monitored and recorded in real time to provide data support for subsequent control and adjustment. At the same time, the cantilever belt pitch angle values ​​at multiple predetermined time points are obtained in order to obtain the changes in the pitch angle information of the cantilever belt within a predetermined time period. Among them, the encoder is a rotation angle measuring device that can convert the rotation angle into a digital signal output. In actual application scenarios, this encoder is usually installed on a cantilever structure, and can monitor the pitch angle changes of the cantilever belt in real time.

[0039] Then, the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points are arranged into a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector according to the time dimension, that is, the time series discrete distribution of the coal flow measurement value and the cantilever belt pitch angle value is converted into a structured vector representation.

[0040] In a specific embodiment of the present application, data preprocessing is performed on the coal flow measurement values ​​at the multiple predetermined time points and the cantilever belt pitch angle values ​​at the multiple predetermined time points to obtain a coal flow measurement value timing input vector and a cantilever belt pitch angle timing input vector, including: arranging the coal flow measurement values ​​at the multiple predetermined time points and the cantilever belt pitch angle values ​​at the multiple predetermined time points into the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector according to the time dimension, respectively.

[0041] In one embodiment of the present application, a timing joint analysis is performed on the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector to obtain a corrected coal flow measurement value timing feature vector, including: performing local neighborhood timing feature extraction on the cantilever belt pitch angle timing input vector to obtain a sequence of cantilever belt pitch angle local timing feature vectors; performing timing feature extraction on the coal flow measurement value timing input vector to obtain a coal flow measurement value timing feature vector; and, embedding the sequence of cantilever belt pitch angle local timing feature vectors and the coal flow measurement value timing feature vector through a feature embedding correction module to obtain the corrected coal flow measurement value timing feature vector.

[0042] Next, the cantilever belt pitch angle time series input vector is vector segmented to obtain a sequence of cantilever belt pitch angle local time series input vectors; and the sequence of cantilever belt pitch angle local time series input vectors is passed through a pitch angle time series feature extractor based on a one-dimensional convolutional layer to obtain a sequence of cantilever belt pitch angle local time series feature vectors. Here, considering that the cantilever belt pitch angle may produce different change characteristics and fluctuation patterns in different time periods, in the technical solution of the present application, vector segmentation is used to guide the model to pay more attention to the change pattern of the cantilever belt pitch angle in different local time periods, so as to understand the dynamic operation law of the cantilever. In addition, the time series local change pattern characteristics of each cantilever belt pitch angle local time series input vector are captured through a one-dimensional convolutional layer.

[0043] In a specific embodiment of the present application, local neighborhood timing feature extraction is performed on the cantilever belt pitch angle timing input vector to obtain a sequence of cantilever belt pitch angle local timing feature vectors, including: vector segmentation of the cantilever belt pitch angle timing input vector to obtain a sequence of cantilever belt pitch angle local timing input vectors; and, passing the sequence of cantilever belt pitch angle local timing input vectors through a pitch angle timing feature extractor based on a one-dimensional convolutional layer to obtain a sequence of cantilever belt pitch angle local timing feature vectors.

[0044] Furthermore, the coal flow measurement value time series input vector is passed through a coal flow measurement time series feature extractor based on a one-dimensional convolutional neural network model to obtain a coal flow measurement value time series feature vector, that is, to capture the time series implicit coal flow measurement chamber change pattern of the coal flow measurement value time series input vector.

[0045] In a specific embodiment of the present application, time series feature extraction is performed on the coal flow measurement value time series input vector to obtain the coal flow measurement value time series feature vector, including: using a deep learning network model to perform feature extraction on the coal flow measurement value time series input vector to obtain the coal flow measurement value time series feature vector.

[0046] Among them, the deep learning network model is a coal flow measurement timing feature extractor based on a one-dimensional convolutional neural network model; wherein, the deep learning network model is used to perform feature extraction on the coal flow measurement value timing input vector to obtain the coal flow measurement value timing feature vector, including: passing the coal flow measurement value timing input vector through the coal flow measurement timing feature extractor based on the one-dimensional convolutional neural network model to obtain the coal flow measurement value timing feature vector.

[0047] Next, the sequence of local time series feature vectors of the cantilever belt pitch angle and the time series feature vectors of the coal flow measurement value are passed through a feature embedding correction module to obtain a corrected coal flow measurement value time series feature vector. It should be understood that as the working state of the bucket wheel machine changes, the pitch angle of the cantilever belt will also change accordingly, causing the distribution of coal on the belt to shift, thereby affecting the center of gravity position of the coal, and the coal flow measurement value of the belt scale will also be inaccurate. In the technical solution of the present application, the local time series dynamic change characteristic information of the cantilever belt pitch angle expressed by the sequence of local time series feature vectors of the cantilever belt pitch angle is used as guiding information to calibrate the description and characterization of the actual coal flow by the coal flow measurement value time series feature vector, so that the corrected coal flow measurement value time series feature vector can better express the true characteristic distribution of the actual coal flow under complex working conditions.

[0048] Furthermore, the sequence of the local time series feature vectors of the cantilever belt pitch angle and the time series feature vector of the coal flow measurement value are passed through a feature embedding correction module to obtain the corrected coal flow measurement value time series feature vector, including: passing the coal flow measurement value time series feature vector through a feature extractor based on a full convolutional network model to obtain a coal flow measurement value feature vector; linearly processing the coal flow measurement value feature vector to obtain a linearly processed coal flow measurement value feature vector; linearly processing the sequence of the local time series feature vectors of the cantilever belt pitch angle to obtain a sequence of the local time series feature vectors of the cantilever belt pitch angle after linear processing. ; Fuse the linearly processed coal flow measurement value feature vector and the linearly processed cantilever belt pitch angle local time series feature vector sequence to obtain the coal flow measurement value-cantilever belt pitch angle linear initial fusion vector; perform one-dimensional convolution processing on the sequence of cantilever belt pitch angle local time series feature vector sequence to obtain the cantilever belt pitch angle local time series neighborhood associated feature vector sequence; and, fuse the cantilever belt pitch angle local time series neighborhood associated feature vector sequence and the coal flow measurement value-cantilever belt pitch angle linear initial fusion vector based on splicing to obtain the corrected coal flow measurement value time series feature vector.

[0049] Furthermore, the time series feature vector of the corrected coal flow measurement value is passed through a decoder to obtain a decoded value, which is used to represent the coal flow correction value at the current time point; and based on the comparison between the decoded value and a predetermined threshold, it is confirmed whether a belt overload warning prompt is generated.

[0050] In a specific embodiment of the present application, based on the corrected coal flow measurement value timing characteristic vector, confirming whether to generate a belt overload warning prompt includes: passing the corrected coal flow measurement value timing characteristic vector through a decoder to obtain a decoded value, wherein the decoded value is used to represent the coal flow correction value at the current time point; and, based on a comparison between the decoded value and a predetermined threshold, confirming whether to generate the belt overload warning prompt.

[0051] In one embodiment of the present application, the automatic control method for coal flow of a bucket wheel excavator further includes a training step: training the coal flow measurement timing feature extractor based on the one-dimensional convolutional neural network model, the pitch angle timing feature extractor based on the one-dimensional convolutional layer, the feature embedding correction module and the decoder; wherein the training step includes: acquiring training data, the training data including training coal flow measurement values ​​at multiple predetermined time points within a predetermined time period collected by a belt scale, training cantilever belt pitch angle values ​​at multiple predetermined time points within the predetermined time period collected by an encoder, and the true value of the coal flow correction value at the current time point; arranging the training coal flow measurement values ​​at the multiple predetermined time points and the training cantilever belt pitch angle values ​​at the multiple predetermined time points into a training coal flow measurement value timing input vector and a training cantilever belt pitch angle timing input vector according to the time dimension respectively; vector segmenting the training cantilever belt pitch angle timing input vector to obtain a sequence of training cantilever belt pitch angle local timing input vectors; and arranging the training coal flow measurement value timing input vector through the training cantilever belt pitch angle local timing input vector based on the training cantilever belt pitch angle local timing input vector. A coal flow measurement timing feature extractor based on a one-dimensional convolutional neural network model is used to obtain a training coal flow measurement value timing feature vector; the sequence of the training cantilever belt pitch angle local timing input vector is passed through the pitch angle timing feature extractor based on the one-dimensional convolutional layer to obtain a sequence of training cantilever belt pitch angle local timing feature vectors; the sequence of the training cantilever belt pitch angle local timing feature vectors and the training coal flow measurement value timing feature vector are passed through the feature embedding correction module to obtain a training correction coal flow measurement value timing feature vector; the training correction coal flow measurement value timing feature vector is passed through a decoder to obtain a decoding loss function value; a predetermined loss function value is calculated between the sequence of the training cantilever belt pitch angle local timing feature vectors and the training coal flow measurement value timing feature vector; and the weighted sum of the decoding loss function value and the predetermined loss function value is used as the loss function value to train the coal flow measurement timing feature extractor based on the one-dimensional convolutional neural network model, the pitch angle timing feature extractor based on the one-dimensional convolutional layer, the feature embedding correction module and the decoder.

[0052] In the technical solution of the present application, the sequence of local time series feature vectors of the training cantilever belt pitch angle expresses the time series distribution of the local time series context association features of the training cantilever belt pitch angle, and the training coal flow measurement value time series feature vector expresses the local time series association features of the training coal flow measurement value. Therefore, when the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector are passed through a feature embedding correction module, the training cantilever belt pitch angle can be corrected based on the local time series association features of the training coal flow measurement value time series feature vector. Therefore, considering the cross-domain feature representation difference between the sequence of local time series feature vectors of the pitch angle of the training cantilever belt and the training coal flow measurement value time series feature vector, if the cross-domain semantic feature sharing between the sequence of local time series feature vectors of the pitch angle of the training cantilever belt and the training coal flow measurement value time series feature vector can be strengthened, the embedded correction expression effect of the training correction coal flow measurement value time series feature vector can be improved.

[0053] That is, considering the feature distribution angles in different source domains and corresponding to different dimensions, the corresponding cross-domain and cross-dimensional sharing of features between the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector may produce feature sharing distribution sparseness under the constraint of feature embedded dynamic distribution. Based on this, the applicant of the present application introduces a predetermined loss function for the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector, which is expressed as: the predetermined loss function value between the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector is calculated by the following optimization formula; wherein, the optimization formula is:

[0054]

[0055] Wherein, V1 is the cascade feature vector obtained by cascading the sequence of the local time series feature vectors of the training cantilever belt pitch angle, and V2 is the time series feature vector of the training coal flow measurement value, V2 T is the transposed vector of the training coal flow measurement value time series 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, Loss represents the predetermined loss function, represents positional subtraction, Represents matrix multiplication.

[0056] Specifically, the strengthening of the shared key features between the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector can be regarded as the distribution information compression of the global feature set. The distribution sparseness 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 local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector. While strengthening the shared key features between the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector, the geometric representation of the sparse but meaningful fusion manifold of the training corrected coal flow measurement value time series feature vector as the fusion feature can be obtained, so as to improve the embedded correction expression effect of the training corrected coal flow measurement value time series feature vector on the sequence of local time series feature vectors of the training cantilever belt pitch angle and the training coal flow measurement value time series feature vector, thereby improving the accuracy of the decoding result obtained by the decoder of the training corrected coal flow measurement value time series feature vector.

[0057] In one embodiment of the present application, the stacking and reclaiming algorithm is the core technology for realizing the intelligentization of the equipment, which is divided into the stacking algorithm and the reclaiming algorithm; the reclaiming algorithm is based on the coal reclaiming plan (giving the coal pile number), under the given initial parameters such as the coal flow control parameters, the trolley stepping parameters, the wheel bucket layer changing thickness parameters, etc., according to the three-dimensional model of the coal pile, to calculate the initial position of the bucket wheel machine for reclaiming coal (tail car position, cantilever rotation angle and pitch angle), left and right rotation angles and other automatic operation parameters of the bucket wheel machine, and pass these parameters to the control system of the bucket wheel machine. When the operator issues a start command, the bucket wheel machine will automatically move to the initial position and start to perform the reclaiming operation.

[0058] The stacking algorithm is to give a coal stacking plan, and then calculate the initial position of the bucket wheel machine's coal stacking (tail car position, cantilever rotation angle and pitch angle), left and right rotation angle and other automatic operation parameters of the bucket wheel machine according to the three-dimensional model of the yard (coal stacking in an empty space or supplementary stacking), and pass the calculated parameters to the on-board control system, and the bucket wheel machine automatically performs stacking operations as required.

[0059] In one embodiment of the present application, protection technology is an indispensable technology for realizing intelligent unmanned stacking and reclaiming equipment. Since the equipment is no longer monitored by people throughout the process after intelligent unmanned operation is realized, obstacles and dangerous factors that are traditionally identified manually must be automatically identified by the equipment. Traditional protection functions are mainly aimed at the protection of the equipment itself, such as limit protection, mechanism interlocking protection, over-temperature protection, over-current protection, etc. After realizing intelligent unmanned operation, on the basis of retaining these protections, more protection is needed for the equipment itself and other equipment and personnel, such as anti-buffer protection, anti-overload protection, protection against collision between personnel and vehicles, and communication interruption protection, etc. The transformation work needs to add distance detection devices, add bucket detection equipment and judgment logic programs, add coal flow detection equipment and algorithms, etc.

[0060] When two bucket wheel machines are operating unmanned at the same time, and there is a possibility that the buckets (or cantilevers) of the two bucket wheel machines may collide in the adjacent coal yard process, it is necessary to give an alarm and automatically stop the operation based on the spatial position judgment of the two bucket wheel machines. In addition, if there is an unmanned automatic operation of a bucket wheel machine and a pusher (coal yard shaping equipment) is running nearby, the safety distance is automatically calculated and judged, and a warning is issued in time or even the operation is stopped. In this case, the pusher needs to be positioned and modified, and the real-time position of the pusher is sent to the unmanned computing server of the bucket wheel machine.

[0061] In summary, an automatic control system and method for coal flow for a bucket wheel excavator based on the embodiment of the present application is explained, which utilizes a belt scale and an encoder deployed on a cantilever belt as data acquisition devices, and utilizes a deep learning algorithm to process and analyze the coal flow measurement values ​​collected by the belt scale and the cantilever belt pitch angle values ​​collected by the encoder, and corrects the actual measurement values ​​of the coal flow using the timing characteristic information of the cantilever belt pitch angle, so that the coal flow measurement values ​​of the belt scale can return to a normal level, and based on this, implements intelligent early warning prompts for belt overloads, so as to avoid equipment damage and safety accidents caused by belt overloads.

[0062] In one embodiment of the present application, Figure 5 FIG. 1 is a block diagram of an automatic coal flow control system for a bucket wheel machine according to an embodiment of the present application. Figure 5As shown, according to the embodiment of the present application, the automatic control system 200 for coal flow of bucket wheel excavator includes: a coal flow measurement value acquisition module 210, which is used to obtain the coal flow measurement values ​​at multiple predetermined time points within a predetermined time period collected by the belt scale; a pitch angle value acquisition module 220, which is used to obtain the cantilever belt pitch angle values ​​at multiple predetermined time points within the predetermined time period collected by the encoder; a data preprocessing module 230, which is used to perform data preprocessing on the coal flow measurement values ​​at the multiple predetermined time points and the cantilever belt pitch angle values ​​at the multiple predetermined time points to obtain the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector; a timing joint analysis module 240, which is used to perform timing joint analysis on the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector to obtain the corrected coal flow measurement value timing feature vector; and a belt overload warning prompt determination module 250, which is used to confirm whether to generate a belt overload warning prompt based on the corrected coal flow measurement value timing feature vector.

[0063] Here, those skilled in the art will appreciate that the specific functions and operations of the various units and modules in the above-mentioned coal flow automatic control system for bucket wheel machine have been described in the above reference. Figures 1 to 4 The description of the automatic coal flow control method for a bucket wheel machine has been described in detail, and therefore, its repeated description will be omitted.

[0064] As described above, the coal flow automatic control system 200 for bucket wheel machine according to the embodiment of the present application can be implemented in various terminal devices, such as a server for automatic control of coal flow of bucket wheel machine, etc. In one example, the coal flow automatic control system 200 for bucket wheel machine according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the coal flow automatic control system 200 for bucket wheel machine can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the coal flow automatic control system 200 for bucket wheel machine can also be one of the many hardware modules of the terminal device.

[0065] Alternatively, in another example, the coal flow automatic control system 200 for bucket wheel excavator and the terminal device may also be separate devices, and the coal flow automatic control system 200 for bucket wheel excavator may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0066] Figure 6 FIG. 1 is a schematic diagram of a method for automatically controlling coal flow of a bucket wheel excavator according to an embodiment of the present application. Figure 6As shown, in this application scenario, first, the coal flow measurement values ​​collected by the belt scale at multiple predetermined time points within a predetermined time period are obtained (for example, Figure 6 and obtaining the cantilever belt pitch angle values ​​at a plurality of predetermined time points within the predetermined time period collected by the encoder (for example, Figure 6 Then, the coal flow measurement value and the cantilever belt pitch angle value are input to a server (for example, Figure 6 In S) shown in the figure, the server can be used for the automatic coal flow control algorithm of the bucket wheel machine to process the coal flow measurement value and the cantilever belt pitch angle value to confirm whether a belt overload warning prompt is generated.

[0067] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0068] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0069] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although multiple example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A method for automatically controlling coal flow for a bucket wheel excavator, characterized in that: include: Obtaining coal flow measurement values ​​collected by the belt scale at a plurality of predetermined time points within a predetermined time period; Acquire the cantilever belt pitch angle values ​​at a plurality of predetermined time points within the predetermined time period collected by the encoder; Performing data preprocessing on the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points to obtain a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector; Performing a time series joint analysis on the coal flow measurement value time series input vector and the cantilever belt pitch angle time series input vector to obtain a corrected coal flow measurement value time series feature vector; as well as Based on the time series characteristic vector of the corrected coal flow measurement value, it is confirmed whether a belt overload warning prompt is generated.

2. The method for automatic coal flow control for a bucket wheel excavator according to claim 1, characterized in that: Data preprocessing is performed on the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points to obtain a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector, including: The coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points are arranged according to the time dimension into the coal flow measurement value timing input vector and the cantilever belt pitch angle timing input vector respectively.

3. The method for automatic coal flow control for a bucket wheel excavator according to claim 2, characterized in that: Performing a time series joint analysis on the coal flow measurement value time series input vector and the cantilever belt pitch angle time series input vector to obtain a corrected coal flow measurement value time series feature vector, including: Performing local neighborhood time series feature extraction on the cantilever belt pitch angle time series input vector to obtain a sequence of cantilever belt pitch angle local time series feature vectors; Performing time series feature extraction on the coal flow measurement value time series input vector to obtain a coal flow measurement value time series feature vector; and The sequence of the local time series feature vectors of the cantilever belt pitch angle and the time series feature vectors of the coal flow measurement value are passed through a feature embedding correction module to obtain the corrected coal flow measurement value time series feature vector.

4. The method for automatic coal flow control for a bucket wheel excavator according to claim 3, characterized in that: The local neighborhood time series feature extraction is performed on the cantilever belt pitch angle time series input vector to obtain a sequence of cantilever belt pitch angle local time series feature vectors, including: Performing vector segmentation on the cantilever belt pitch angle timing input vector to obtain a sequence of cantilever belt pitch angle local timing input vectors; and The sequence of the local time series input vectors of the cantilever belt pitch angle is passed through a pitch angle time series feature extractor based on a one-dimensional convolutional layer to obtain a sequence of the local time series feature vectors of the cantilever belt pitch angle.

5. The method for automatic coal flow control for a bucket wheel excavator according to claim 4, characterized in that: Extracting time series features from the coal flow measurement value time series input vector to obtain a coal flow measurement value time series feature vector includes: A deep learning network model is used to perform feature extraction on the coal flow measurement value time series input vector to obtain the coal flow measurement value time series feature vector.

6. The method for automatic coal flow control for a bucket wheel excavator according to claim 5, characterized in that: The deep learning network model is a coal flow measurement time series feature extractor based on a one-dimensional convolutional neural network model; Among them, using a deep learning network model to extract features from the coal flow measurement value time series input vector to obtain the coal flow measurement value time series feature vector includes: The coal flow measurement value time series input vector is passed through the coal flow measurement time series feature extractor based on the one-dimensional convolutional neural network model to obtain the coal flow measurement value time series feature vector.

7. The method for automatic coal flow control for a bucket wheel excavator according to claim 6, characterized in that: The sequence of the local time series feature vectors of the cantilever belt pitch angle and the time series feature vectors of the coal flow measurement value are passed through a feature embedding correction module to obtain the corrected coal flow measurement value time series feature vector, including: The coal flow measurement value time series feature vector is passed through a feature extractor based on a fully convolutional network model to obtain a coal flow measurement value feature vector; Performing linear processing on the coal flow measurement value characteristic vector to obtain a linearly processed coal flow measurement value characteristic vector; Performing linear processing on the sequence of local time series feature vectors of the cantilever belt pitch angle to obtain a sequence of local time series feature vectors of the cantilever belt pitch angle after linear processing; Fusion of the linearly processed coal flow measurement value feature vector and the linearly processed cantilever belt pitch angle local time series feature vector sequence to obtain a coal flow measurement value-cantilever belt pitch angle linear initial fusion vector; Performing one-dimensional convolution processing on the sequence of local time series feature vectors of the cantilever belt pitch angle to obtain a sequence of local time series neighborhood correlation feature vectors of the cantilever belt pitch angle; and The sequence of the local time series neighborhood correlation feature vectors of the cantilever belt pitch angle and the linear initial fusion vector of the coal flow measurement value-cantilever belt pitch angle are fused based on the splicing method to obtain the corrected coal flow measurement value time series feature vector.

8. The method for automatic coal flow control for a bucket wheel excavator according to claim 7, characterized in that: Based on the time series characteristic vector of the corrected coal flow measurement value, confirming whether a belt overload warning prompt is generated includes: Passing the corrected coal flow measurement value time series feature vector through a decoder to obtain a decoded value, wherein the decoded value is used to represent the coal flow correction value at the current time point; and Based on the comparison between the decoded value and a predetermined threshold, it is determined whether the belt overload warning prompt is generated.

9. The method for automatic coal flow control for a bucket wheel excavator according to claim 8, characterized in that: The method further includes a training step of: training the coal flow measurement time series feature extractor based on the one-dimensional convolutional neural network model, the pitch angle time series feature extractor based on the one-dimensional convolutional layer, the feature embedding correction module and the decoder; Wherein, the training step includes: Acquire training data, the training data including training coal flow measurement values ​​at multiple predetermined time points within a predetermined time period collected by a belt scale, training cantilever belt pitch angle values ​​at multiple predetermined time points within the predetermined time period collected by an encoder, and a true value of a coal flow correction value at a current time point; Arrange the training coal flow measurement values ​​at the plurality of predetermined time points and the training cantilever belt pitch angle values ​​at the plurality of predetermined time points into a training coal flow measurement value time series input vector and a training cantilever belt pitch angle time series input vector according to the time dimension respectively; Performing vector segmentation on the training cantilever belt pitch angle time series input vector to obtain a sequence of training cantilever belt pitch angle local time series input vectors; Passing the training coal flow measurement value time series input vector through the coal flow measurement time series feature extractor based on the one-dimensional convolutional neural network model to obtain a training coal flow measurement value time series feature vector; Passing the sequence of the training cantilever belt pitch angle local time series input vectors through the pitch angle time series feature extractor based on the one-dimensional convolution layer to obtain a sequence of the training cantilever belt pitch angle local time series feature vectors; The sequence of the training cantilever belt pitch angle local time series feature vectors and the training coal flow measurement value time series feature vectors are passed through the feature embedding correction module to obtain the training correction coal flow measurement value time series feature vector; Passing the training corrected coal flow measurement value time series feature vector through a decoder to obtain a decoding loss function value; Calculating a predetermined loss function value between the sequence of the training cantilever belt pitch angle local time series feature vectors and the training coal flow measurement value time series feature vectors; and The weighted sum of the decoding loss function value and the predetermined loss function value is used as the loss function value to train the coal flow measurement timing feature extractor based on the one-dimensional convolutional neural network model, the pitch angle timing feature extractor based on the one-dimensional convolutional layer, the feature embedding correction module and the decoder.

10. A coal flow automatic control system for a bucket wheel excavator, characterized in that: include: A coal flow measurement value acquisition module, used to acquire coal flow measurement values ​​collected by a belt scale at a plurality of predetermined time points within a predetermined time period; A pitch angle value acquisition module, used to acquire the cantilever belt pitch angle values ​​at a plurality of predetermined time points within the predetermined time period collected by the encoder; A data preprocessing module, used for performing data preprocessing on the coal flow measurement values ​​at the plurality of predetermined time points and the cantilever belt pitch angle values ​​at the plurality of predetermined time points to obtain a coal flow measurement value time series input vector and a cantilever belt pitch angle time series input vector; A time series joint analysis module, used for performing a time series joint analysis on the coal flow measurement value time series input vector and the cantilever belt pitch angle time series input vector to obtain a corrected coal flow measurement value time series feature vector; as well as The belt overload warning prompt determination module is used to confirm whether to generate a belt overload warning prompt based on the time series feature vector of the corrected coal flow measurement value.

Images