Coal blockage shutdown control system for coal drop pipe

Abstract

The invention provides a coal blockage shutdown control system for a coal drop pipe, and relates to the technical field of coal conveying, and the system comprises a coal blockage detection module which is used for detecting coal blockage condition data in the coal drop pipe; the signal processing module is used for sequentially carrying out filtering and wavelet transformation processing on the coal blockage condition data; the double-shutdown judgment module is used for carrying out coal blockage analysis on the coal blockage condition data and the wavelet transformation result based on a support vector machine model and correspondingly generating a first judgment result and a second judgment result; the shutdown control module is used for comprehensively analyzing the first judgment result and the second judgment result, generating a corresponding control instruction and controlling actions of related equipment of the coal drop pipe; the alarm module receives the control instruction and provides a corresponding alarm action; by adopting a multi-sensor fusion technology, an advanced signal processing method, double-model judgment and a redundancy control mechanism, the accuracy of coal blockage monitoring of the coal drop pipe, the effectiveness of signal processing, the reliability of shutdown control and the completeness of an alarm mechanism can be effectively improved.

Description

Technical Field

[0001] The invention relates to the technical field of coal transportation, and in particular to a coal blocking shutdown control system for a coal drop pipe. Background Art

[0002] With the continuous development of the coal industry, the automation control technology of coal transportation equipment has also gone through several stages of development. In the early days, coal transportation mainly relied on manual operation and simple mechanical control, which had problems such as low efficiency, high labor intensity and many safety hazards. With the rise of electronic technology and automation technology, coal transportation equipment began to introduce sensor technology for simple monitoring, such as installing speed sensors and deviation sensors on conveyor belts and other parts to achieve basic monitoring of the equipment's operating status and timely discover some common faults.

[0003] However, in terms of coal pipe blockage monitoring, the existing technology often uses a single sensor or a simple monitoring method. For example, when only a pressure sensor is used, when the pressure in the coal pipe fluctuates, it may be due to a variety of reasons (such as instantaneous impact of materials, etc.), and it is difficult to accurately determine whether the pressure change is caused by coal blockage or normal operation changes. In terms of shutdown control, the existing technology may have the problem of single control logic. For example, when a possible coal blockage signal is detected, the shutdown is directly triggered without considering the accuracy and reliability of the signal, which may cause an erroneous shutdown and affect the continuity of coal transportation.

[0004] Therefore, there is an urgent need in the art for a coal-blocking shutdown control system for a coal drop pipe to solve the above-mentioned problem. Summary of the invention

[0005] The present invention provides a coal blockage shutdown control system for a coal chute. In view of the problems existing in the above-mentioned prior art, the accuracy of coal chute blockage monitoring, the effectiveness of signal processing, the reliability of shutdown control and the perfection of the alarm mechanism are effectively improved by adopting multi-sensor fusion technology, advanced signal processing methods, dual model judgment and redundant control mechanism.

[0006] The present invention provides a coal blocking shutdown control system for a coal drop pipe, comprising:

[0007] A coal blocking detection module is arranged in the coal dropping pipe and is used to detect the coal blocking data in the coal dropping pipe;

[0008] A signal processing module, which is connected to the coal blocking detection module and is used to perform filtering and wavelet transformation processing on the coal blocking situation data in sequence to generate a wavelet transformation result;

[0009] A dual shutdown judgment module, which is connected to the coal blocking detection module and the signal processing module, performs coal blocking analysis on the coal blocking situation data and the wavelet transform result based on the pre-trained support vector machine model, and generates a first judgment result and a second judgment result accordingly;

[0010] A shutdown control module, which is connected to the dual shutdown judgment module, comprehensively analyzes the first judgment result and the second judgment result based on a preset logic, generates a corresponding control instruction, and controls the action of the coal drop pipe related equipment according to the control instruction;

[0011] An alarm module is connected to the shutdown control module and is used to synchronously receive the control instruction and provide a corresponding alarm action.

[0012] According to a coal blocking shutdown control system for a coal drop pipe provided by the present invention, the coal blocking detection module comprises:

[0013] The pressure sensing unit is evenly distributed on the inner wall of the coal drop pipe and is used to detect the pressure change in the current coal drop pipe. The detection value of the pressure sensing unit is P (t) ;

[0014] The ultrasonic sensor unit is used to detect the material height in the coal drop pipe. The detection value of the ultrasonic sensor unit is H (t) ;

[0015] The laser distance measuring unit is arranged at the top of the coal dropping pipe and is used to detect the distance from the material surface in the coal dropping pipe to the sensor. The detection value of the laser distance measuring unit is L (t) ;

[0016] A signal output unit connected to the pressure sensing unit, the ultrasonic sensing unit and the laser ranging unit, and used to take the pressure, material height and distance as three components of a vector in a three-dimensional space;

[0017] By formula Calculate the comprehensive value after integrating the three components, that is, the coal blocking data F (t) .

[0018] According to a coal blocking shutdown control system for a coal drop pipe provided by the present invention, the signal processing module comprises:

[0019] The adaptive filtering unit is connected to the signal output unit and is used to receive the coal blocking situation data F (t) , and use adaptive filtering algorithm to remove noise interference to obtain the filtered signal Y (n) ;

[0020] Wavelet transform unit; it is connected to the adaptive filtering unit and is used to filter the signal Y through the wavelet transform model (n)Perform wavelet transform and get the wavelet transform result W X(a，b) ; The wavelet transform model is:

[0021]

[0022] Among them, Y (n) is the filtered signal at the current time t; a is the scale parameter, which is used to determine the degree of expansion and contraction of the wavelet transform result. The appropriate scale parameter range is determined through experiments or experience, and then the discretization value is taken within this range; is the translation parameter, which is used to determine the position of the wavelet transform result on the time axis, which is determined according to the length of the signal and the scale parameter; is the mother wavelet function, and one of the Haar wavelet, Daubechies wavelet, and Morlet wavelet is selected based on the signal characteristics.

[0023] According to a coal-blocking shutdown control system for a coal drop pipe provided by the present invention, in the adaptive filtering unit, an adaptive filtering algorithm is used to remove noise interference to obtain a filtering signal Y (n) The process includes:

[0024] Preset the initial weight coefficient W (0) , step size factor μ and expected signal d (n) ;

[0025] The initial weight coefficient W (0) The value is determined based on experience; the step factor μ is used to control the update speed of the weight coefficient and is determined through experiments or experience; the fused signal after a period of delay is used as the expected signal d (n) , or determine an ideal signal based on historical data;

[0026] Receive coal blocking data F (t) As the input signal X (n) ;

[0027] Based on the formula (n) =d (n) -Y (n) Calculate the error signal e (n) , where Y (n) =H (z) *X (n) , Error signal e (n) Used to reflect the difference between the expected signal and the filter output signal;

[0028] According to the formula W (n+1) =W (n) +μe (n) *X (n) Update the weight coefficients so that the output of the filter gradually approaches the expected signal;

[0029] Calculate the filter output signal Y using the updated weight coefficients (n+1) =H (z) *×X (n+1) ;in

[0030] The above process is repeated to continuously perform error calculation, weight coefficient update and output signal calculation until the preset number of iterations is reached.

[0031] According to a coal blocking shutdown control system for a coal drop pipe provided by the present invention, the double shutdown judgment module comprises:

[0032] A machine learning unit, which establishes a first historical training set M=(F (i) ,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the first historical training set M is substituted into the support vector machine model for training to generate a first judgment model;

[0033] Based on the historical wavelet transform results and the actual coal blocking situation, the second historical training set N = (W X(a，b) i,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the second historical training set N is substituted into the support vector machine model for training to generate a second judgment model;

[0034] The result output unit is connected to the machine learning unit and is used to output the current coal blocking situation data F (t) Substitute into the first judgment model to generate the label result Y t , corresponding to the first judgment result S1(t); the current wavelet transform result W X(a，b) Substitute into the second judgment model to generate the label result Y t , and a second judgment result S2(t) is generated accordingly.

[0035] According to a coal blocking shutdown control system for a coal drop pipe provided by the present invention, the shutdown control module comprises:

[0036] A redundancy control unit connected to the result output unit, for receiving the first judgment result S1(t) and the second judgment result S2(t), and performing redundancy control by majority voting; that is, when S1(t)=S2(t), the output redundancy control result is S(t)=S1(t); when S1(t)≠S2(t), S2(t) is defined to take precedence over S1(t), and the output redundancy control result is S(t)=S2(t);

[0037] An instruction generating unit, connected to the redundant control unit, and generating corresponding control instructions based on the redundant control result S(t);

[0038] The shutdown control terminal is connected to the instruction generation unit, and is used to receive the control instruction and perform a shutdown operation corresponding to the content of the control instruction.

[0039] According to a coal-blocking shutdown control system for a coal drop pipe provided by the present invention, the alarm module comprises:

[0040] A mobile device, which is wirelessly connected to the instruction generation unit and is used to display the control instruction corresponding to the current redundant control result in real time;

[0041] The sound and light alarm unit is connected to the instruction generating unit and is arranged at the corresponding coal dropping pipe, and is used to perform sound and light alarm actions according to the control instructions.

[0042] According to a coal-drop pipe blockage shutdown control system provided by the present invention, the support vector machine model in the machine learning unit uses a cross-validation method to select the optimal parameter combination during the training process.

[0043] Compared with the prior art, the beneficial effects of this application are:

[0044] This application obtains information from multiple angles through multi-sensor fusion, which is more comprehensive and accurate than single sensor monitoring; it adopts dual-model judgment, analyzes from macro and micro perspectives, complements and verifies each other, and further improves the accuracy of judgment; this application adopts an adaptive filtering algorithm to remove noise interference, and uses wavelet transform to perform multi-scale analysis of signals, which can better extract useful features and overcome the shortcomings of signal processing methods in the prior art; this application adopts a redundant control mechanism, and controls the two judgment results through majority voting to avoid false shutdowns caused by misjudgment of a single model, while still ensuring reliable shutdown when a control system fails.

[0045] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0046] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0048] Figure 1 It is a structural schematic diagram of a coal blocking shutdown control system for a coal drop pipe provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0050] Embodiment 1:

[0051] The embodiment of the present invention provides a coal blocking shutdown control system for a coal drop pipe, please refer to Figure 1 ,include:

[0052] A coal blocking detection module is arranged in the coal dropping pipe and is used to detect the coal blocking data in the coal dropping pipe;

[0053] A signal processing module, which is connected to the coal blocking detection module and is used to filter and wavelet transform the coal blocking situation data in sequence to generate a wavelet transform result;

[0054] A dual shutdown judgment module, which is connected to the coal blocking detection module and the signal processing module, performs coal blocking analysis on the coal blocking situation data and the wavelet transform result based on the pre-trained support vector machine model, and generates a first judgment result and a second judgment result accordingly;

[0055] A shutdown control module is connected to the dual shutdown judgment module, and comprehensively analyzes the first judgment result and the second judgment result based on a preset logic, generates a corresponding control instruction, and controls the action of the coal drop pipe related equipment according to the control instruction;

[0056] The alarm module is connected to the shutdown control module and is used to synchronously receive control instructions and provide corresponding alarm actions.

[0057] The principle of this embodiment is: the coal blockage detection module is located inside the coal drop pipe, and monitors the flow of coal in the coal drop pipe in real time through sensors and other equipment, and collects data related to coal blockage; the signal processing module receives data from the coal blockage detection module, and filters these raw data to remove noise, and then uses wavelet transform technology to further process the data to extract feature information that is more conducive to analysis; the double shutdown judgment module adopts the support vector machine (SVM) algorithm, and independently analyzes the processed coal blockage data (that is, the data directly obtained from the coal blockage detection module) and the wavelet transform results based on the trained model; such a dual judgment mechanism can improve the judgment accuracy; the shutdown control module comprehensively evaluates whether the current situation requires stopping the coal drop operation according to the preset logical rules based on the two judgment results provided by the double shutdown judgment module, and issues corresponding control instructions to relevant mechanical equipment for execution; once the alarm module decides to take shutdown measures, the module will synchronously receive the control instructions and start the alarm system to notify the on-site staff to take timely response measures.

[0058] The beneficial effects of this embodiment are: by quickly and accurately identifying and responding to coal blockage, safety accidents that may be caused by failure to discover in time are effectively avoided; the use of advanced signal processing technology and machine learning methods improves the system's ability to identify abnormal conditions under complex working conditions, ensuring the stable operation of the system; it is able to detect and deal with coal blockage problems in a timely manner, reducing unnecessary downtime and the economic losses caused by it; the entire process is highly automated, reducing the possibility of interference from human factors, and also reducing the workload of operators.

[0059] In order to further optimize the above embodiment, the coal blocking detection module includes:

[0060] The pressure sensing unit is evenly distributed on the inner wall of the coal drop pipe and is used to detect the pressure change in the current coal drop pipe. The detection value of the pressure sensing unit is P (t) ;

[0061] The ultrasonic sensor unit is used to detect the material height in the coal drop pipe. The detection value of the ultrasonic sensor unit is H (t) ;

[0062] The laser distance measuring unit is set at the top of the coal drop pipe to detect the distance from the material surface in the coal drop pipe to the sensor. The detection value of the laser distance measuring unit is L (t) ;

[0063] A signal output unit, which is connected to the pressure sensing unit, the ultrasonic sensing unit, and the laser ranging unit, and is used to take the pressure, the material height, and the distance as three components of a vector in a three-dimensional space;

[0064] By formula Calculate the comprehensive value after integrating the three components, that is, the coal blocking data F (t) .

[0065] It should be noted that the pressure sensing unit is based on the physical relationship between pressure and material accumulation. When coal is blocked in the coal drop pipe, material accumulation will cause pressure changes. The pressure sensor senses this pressure change and converts it into an electrical signal for output. It can use a piezoresistive pressure sensor, a capacitive pressure sensor, etc. These sensors have different working principles and characteristics and can be selected according to actual needs and environmental conditions. For example, a piezoresistive pressure sensor uses the piezoresistive effect of semiconductor materials. When subjected to pressure, its resistance value changes, and the pressure is determined by measuring the resistance change.

[0066] The ultrasonic sensing unit uses the propagation characteristics of ultrasonic waves in the medium. The ultrasonic sensor emits ultrasonic pulses into the coal drop pipe. When encountering the surface of the material, the ultrasonic wave will be reflected back. By measuring the time interval from the emission to the reception of the ultrasonic wave, and according to the propagation speed of the ultrasonic wave in the medium, the height of the material can be calculated. Common ultrasonic sensors include integrated transceiver and split transceiver. The integrated transceiver sensor integrates the transmission and reception functions into one probe, with a simple structure, and is suitable for some occasions where the accuracy requirements are not particularly high. The split transceiver sensor sets the transmitting probe and the receiving probe separately, and the measurement accuracy can be improved by adjusting the position and angle of the probe, which is suitable for more complex measurement environments.

[0067] The laser distance measuring unit is based on the reflection principle of laser. The laser distance meter emits a laser beam into the coal drop pipe. When the laser beam hits the surface of the material, it will be reflected back. By measuring the time difference between laser emission and reflection and combining it with the speed of light, the distance from the material surface to the sensor can be calculated. It can use pulse laser distance measurement or phase laser distance measurement. The pulse laser distance measurement method calculates the distance by measuring the flight time of the laser pulse, and has the characteristics of long measurement distance and high accuracy. The phase laser distance measurement method calculates the distance by measuring the phase difference between the laser emission and reflection signals, and has higher accuracy, but the measurement distance is relatively short. The appropriate method can be selected according to the actual size of the coal drop pipe and the measurement accuracy requirements.

[0068] The fusion of pressure, material height and distance as three components of a vector in three-dimensional space is based on their interrelated and complementary characteristics in reflecting coal blockage. When coal blockage occurs, these three parameters usually change at the same time, and the fused comprehensive value can more comprehensively reflect the coal blockage. The principle of the formula is similar to the modulus length calculation formula of the space vector. The square root of the sum of the squares of the three components is taken to obtain a value that comprehensively reflects the changes in the three parameters. When any of the parameters changes significantly, the fused comprehensive value will be significantly affected, so that the occurrence of coal blockage can be detected in time.

[0069] In order to further optimize the above embodiment, the signal processing module includes:

[0070] The adaptive filtering unit is connected to the signal output unit and is used to receive the coal blocking data F (t) , and use adaptive filtering algorithm to remove noise interference to obtain the filtered signal Y (n) ;

[0071] In the adaptive filtering unit, an adaptive filtering algorithm is used to remove noise interference and obtain the filtered signal Y (n) The process includes:

[0072] Preset the initial weight coefficient W (0) , step size factor μ and expected signal d (n) ;

[0073] Initial weight coefficient W (0) The value of is determined by experience; the step factor μ is used to control the update speed of the weight coefficient, which is determined by experiment or experience; the fused signal after a period of delay is taken as the expected signal d (n) , or determine an ideal signal based on historical data;

[0074] Receive coal blocking data F (t) As the input signal X (n) ;

[0075] Based on the formula (n) =d (n) -Y (n) Calculate the error signal e (n) , where Y (n) =H (z) *X (n) , Error signal e (n) Used to reflect the difference between the expected signal and the filter output signal;

[0076] According to the formula W (n+1) =W (n) +μe (n) *X (n)Update the weight coefficients so that the output of the filter gradually approaches the expected signal;

[0077] Calculate the filter output signal Y using the updated weight coefficients (n+1) =H (z) *×X (n+1) ;in

[0078] Repeat the above process, continuously perform error calculation, weight coefficient update and output signal calculation until the preset number of iterations is reached;

[0079] The wavelet transform unit is connected to the adaptive filter unit and is used to filter the signal Y through the wavelet transform model. (n) Perform wavelet transform and get the wavelet transform result W X(a，b) ; The wavelet transform model is:

[0080]

[0081] Among them, Y (n) is the filtered signal at the current time t; a is the scale parameter, which is used to determine the degree of expansion and contraction of the wavelet transform result. The appropriate scale parameter range is determined through experiments or experience, and then the discretization value is taken within this range; is the translation parameter, which is used to determine the position of the wavelet transform result on the time axis, which is determined according to the length of the signal and the scale parameter; is the mother wavelet function, and one of the Haar wavelet, Daubechies wavelet, and Morlet wavelet is selected based on the signal characteristics.

[0082] It should be noted that the purpose of adaptive filtering is to automatically adjust the filter parameters according to the characteristics of the input signal to achieve the best filtering effect; in this system, the adaptive filtering unit receives coal blocking data (including information after the fusion of pressure, material height and distance), which may be interfered by various noises, such as noise generated by electromagnetic interference, environmental vibration, etc.; by continuously adjusting the weight coefficient of the filter, the output signal of the filter is made as close to the expected signal as possible, thereby removing noise interference; the expected signal can be a fused signal after a period of delay, or an ideal signal determined based on historical data and experience;

[0083] Wavelet transform is a time-frequency analysis method that can decompose the signal into components of different scales and frequencies, thereby extracting the local features and detailed information of the signal. In this system, the wavelet transform unit processes the signal after adaptive filtering to further explore the features related to coal blocking in the signal.

[0084] Different scale parameters correspond to the analysis of different frequency ranges; larger scale parameters can capture the low-frequency components and overall trends in the signal, and smaller scale parameters can capture the high-frequency components and local details in the signal; the translation parameter determines the position of the wavelet transform result on the time axis. By changing the value, the wavelet transform can analyze the signal at different time positions;

[0085] The selection of the mother wavelet function is based on the characteristics of the signal; different mother wavelet functions have different shapes and properties, and are suitable for different types of signal analysis; for example, the Haar wavelet is a simple wavelet function with good localization characteristics, suitable for analyzing signals with mutation points; the Daubechies wavelet has better smoothness and multi-resolution analysis capabilities, suitable for analyzing complex signals; the Morlet wavelet is suitable for analyzing signals with periodic or oscillatory characteristics;

[0086] For example, if the flow of materials in the coal drop pipe suddenly starts to block or dredge, it may cause sudden changes in signals such as pressure, material height and distance. For example, when a large piece of coal is suddenly stuck in a certain place in the coal drop pipe, the local pressure increases instantly, the material height changes suddenly near the position, and the distance detected by the laser ranging sensor may also change suddenly. In this case, the good localization characteristics of the Haar wavelet and its ability to detect mutation points make it a suitable choice. It can quickly locate the location of these signal mutations, which helps to detect such sudden coal blockages in a timely manner.

[0087] When considering more complex signal conditions in the coal drop pipe, such as during normal operation, the flow of materials is a complex dynamic process. Signals such as pressure, material height and distance will be affected by many factors, including the particle size distribution of the material, the fluctuation of the coal drop speed, etc. For example, when coal particles of different sizes flow together, it may cause complex fluctuations in the pressure signal and irregular changes in the material height. The compact support and multi-resolution analysis capabilities of the Daubechies wavelet can better process such complex signals. It can capture the overall trend of the signal while retaining the local characteristics of the signal, which helps to more accurately analyze the possibility of coal blockage in such complex situations.

[0088] If the material flow in the coal drop pipe presents certain periodic characteristics, for example, due to the periodic feeding action of the coal feeder, the material flow in the coal drop pipe presents periodic pressure changes, material height changes, etc.; in this case, the good frequency localization characteristics and periodic matching ability of Morlet wavelet make it a suitable choice; it can better extract characteristic information related to the period, which helps to judge whether this periodic change is related to coal blockage, or whether it will cause coal blockage;

[0089] Therefore, the specific choice should be made based on the actual situation and will not be elaborated in detail here.

[0090] In order to further optimize the above embodiment, the double shutdown judgment module includes:

[0091] A machine learning unit, which establishes a first historical training set M=(F (i) ,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the first historical training set M is substituted into the support vector machine model for training to generate a first judgment model;

[0092] Based on the historical wavelet transform results and the actual coal blocking situation, the second historical training set N = (W X(a，b) i,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the second historical training set N is substituted into the support vector machine model for training to generate a second judgment model;

[0093] The result output unit is connected to the machine learning unit and is used to output the current coal blocking situation data F (t) Substitute into the first judgment model to generate the label result Y t , corresponding to the first judgment result S1(t); the current wavelet transform result W X(a，b) Substitute into the second judgment model to generate the label result Y t , and a second judgment result S2(t) is generated accordingly.

[0094] It should be noted that the coal blocking data is a comprehensive value obtained by fusing the pressure, material height and distance information, which reflects the overall state information in the coal drop pipe; the first judgment model trained based on this data can judge the coal blocking situation from a macro perspective;

[0095] The wavelet transform result is obtained by multi-scale analysis of the filtered signal, which focuses more on extracting the local features and detail information of the signal. The second judgment model trained based on the wavelet transform result can capture the features related to coal blockage from a microscopic perspective.

[0096] A single model cannot take into account both overall and local information at the same time. For example, if only a model based on coal blockage data is used, some local subtle changes will be ignored, and these changes are important early signs of coal blockage. On the contrary, if only a model based on wavelet transform results is relied upon, misjudgment may occur due to lack of overall information, such as misjudging some normal local fluctuations as coal blockage.

[0097] The two models analyze the coal blockage situation from different angles, and their judgment results can complement and verify each other; when both models judge that there is coal blockage, the reliability of the result is higher; when the judgments of the two models are inconsistent, the reasons can be further analyzed to avoid misjudgment of a single model; under different working conditions and environments, the situation in the coal drop pipe is complex and changeable; the dual model can better adapt to such changes and improve the robustness of the system in various situations; for example, under certain special material properties or coal drop speeds, one model is affected, but the other model can still make accurate judgments, thereby ensuring the overall reliability of the system.

[0098] In order to further optimize the above embodiment, the shutdown control module includes:

[0099] A redundancy control unit connected to the result output unit, for receiving the first judgment result S1(t) and the second judgment result S2(t), and adopting a majority voting method for redundancy control; that is, when S1(t)=S2(t), the output redundancy control result is S(t)=S1(t); when S1(t)≠S2(t), S2(t) is defined to take precedence over S1(t), and the output redundancy control result is S(t)=S2(t);

[0100] An instruction generating unit, which is connected to the redundant control unit and generates corresponding control instructions based on the redundant control result S(t);

[0101] The shutdown control terminal is connected to the instruction generation unit, and is used to receive the control instruction and perform a shutdown operation corresponding to the content of the control instruction.

[0102] It should be noted that the wavelet transform result can extract the local features and detailed information of the signal. In some complex and local changes that have a greater impact on the judgment of coal blockage, the second judgment model may be more able to accurately capture these key information, and thus be given priority when it is inconsistent with the first judgment model based on the overall fusion data; in a complex industrial environment, the situation in the coal drop pipe is complex and changeable, and there are various interference factors. The priority selection of S2(t) can use its more accurate local feature judgment ability when the two models are inconsistent, improve the accuracy of the system's overall judgment of coal blockage, and reduce the occurrence of misjudgment.

[0103] In order to further optimize the above embodiment, the alarm module includes:

[0104] A mobile device is wirelessly connected to the instruction generation unit and is used to display the control instruction corresponding to the current redundant control result in real time;

[0105] An audible and visual alarm unit, which is connected to the instruction generating unit and is arranged at the corresponding coal dropping pipe, and is used to perform audible and visual alarm actions according to the control instructions;

[0106] Furthermore, during the training process of the support vector machine model in the machine learning unit, a cross-validation method is used to select the optimal parameter combination.

[0107] It should be noted that the support vector machine is a binary classification model. Its basic idea is to find an optimal hyperplane in the feature space, separate samples of different categories as much as possible, and maximize the distance from this hyperplane to the nearest sample point (called support vector) in the two categories. This distance is called margin. For linearly separable data, the optimal hyperplane can be found by solving a quadratic programming problem. For linearly inseparable data, the original data can be mapped to a higher-dimensional feature space by introducing a kernel function, where the data may become linearly separable, and then the optimal hyperplane can be found. This system uses this algorithm as a model for training, and determines on which side of the hyperplane the current wavelet transform result and the current coal blocking data are located, and then determines the actual coal blocking situation.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A coal blocking shutdown control system for a coal drop pipe, characterized in that: include: A coal blocking detection module is arranged in the coal dropping pipe and is used to detect the coal blocking data in the coal dropping pipe; A signal processing module, which is connected to the coal blocking detection module and is used to perform filtering and wavelet transformation processing on the coal blocking situation data in sequence to generate a wavelet transformation result; A dual shutdown judgment module, which is connected to the coal blocking detection module and the signal processing module, performs coal blocking analysis on the coal blocking situation data and the wavelet transform result based on the pre-trained support vector machine model, and generates a first judgment result and a second judgment result accordingly; A shutdown control module, which is connected to the dual shutdown judgment module, comprehensively analyzes the first judgment result and the second judgment result based on a preset logic, generates a corresponding control instruction, and controls the action of the coal drop pipe related equipment according to the control instruction; An alarm module is connected to the shutdown control module and is used to synchronously receive the control instruction and provide a corresponding alarm action.

2. A coal blocking shutdown control system for a coal drop pipe according to claim 1, characterized in that: The coal blocking detection module comprises: The pressure sensing unit is evenly distributed on the inner wall of the coal drop pipe and is used to detect the pressure change in the current coal drop pipe. The detection value of the pressure sensing unit is P (t) ; The ultrasonic sensor unit is used to detect the material height in the coal drop pipe. The detection value of the ultrasonic sensor unit is H (t) ; The laser distance measuring unit is arranged at the top of the coal dropping pipe and is used to detect the distance from the material surface in the coal dropping pipe to the sensor. The detection value of the laser distance measuring unit is L (t) ; A signal output unit connected to the pressure sensing unit, the ultrasonic sensing unit and the laser ranging unit, and used to take the pressure, material height and distance as three components of a vector in a three-dimensional space; By formula Calculate the comprehensive value after integrating the three components, that is, the coal blocking data F (t) .

3. A coal blocking shutdown control system for a coal drop pipe according to claim 2, characterized in that: The signal processing module comprises: The adaptive filtering unit is connected to the signal output unit and is used to receive the coal blocking situation data F (t) , and use adaptive filtering algorithm to remove noise interference to obtain the filtered signal Y (n) ; Wavelet transform unit; it is connected to the adaptive filtering unit and is used to filter the signal Y through the wavelet transform model (n) Perform wavelet transform and get the wavelet transform result W X(a , b) ; The wavelet transform model is: Among them, Y (n) is the filtered signal at the current time t; a is the scale parameter, which is used to determine the degree of expansion and contraction of the wavelet transform result. The appropriate scale parameter range is determined through experiments or experience, and then the discretization value is taken within this range; is the translation parameter, which is used to determine the position of the wavelet transform result on the time axis, which is determined according to the length of the signal and the scale parameter; is the mother wavelet function, and one of the Haar wavelet, Daubechies wavelet, and Morlet wavelet is selected based on the signal characteristics.

4. A coal blocking shutdown control system for a coal drop pipe according to claim 3, characterized in that: In the adaptive filtering unit, an adaptive filtering algorithm is used to remove noise interference to obtain a filtered signal Y (n) The process includes: Preset the initial weight coefficient W (0) , step size factor μ and expected signal d (n) ; The initial weight coefficient W (0) The value is determined based on experience; the step factor μ is used to control the update speed of the weight coefficient and is determined through experiments or experience; the fused signal after a period of delay is used as the expected signal d (n) , or determine an ideal signal based on historical data; Receive coal blocking data F (t) As the input signal X (n) ; Based on the formula (n) =d (n) -Y (n) Calculate the error signal e (n) , where Y (n) =H (z) *X (n) , Error signal e (n) Used to reflect the difference between the expected signal and the filter output signal; According to the formula W (n+1) =W (n) +μe (n) *X (n) Update the weight coefficients so that the output of the filter gradually approaches the expected signal; Calculate the filter output signal Y using the updated weight coefficients (n+1) =H (z) *×X (n+1) ;in The above process is repeated to continuously perform error calculation, weight coefficient update and output signal calculation until the preset number of iterations is reached.

5. A coal blocking shutdown control system for a coal drop pipe according to claim 4, characterized in that: The dual shutdown judgment module comprises: A machine learning unit, which establishes a first historical training set M=(F (i) ,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the first historical training set M is substituted into the support vector machine model for training to generate a first judgment model; Based on the historical wavelet transform results and the actual coal blocking situation, the second historical training set N = (W X(a，b) i,y (i) ), where Y i ∈{-1, 1} is a category label, representing whether there is coal blocking; the second historical training set N is substituted into the support vector machine model for training to generate a second judgment model; The result output unit is connected to the machine learning unit and is used to output the current coal blocking situation data F (t) Substitute into the first judgment model to generate the label result Y t , corresponding to the first judgment result S1(t); the current wavelet transform result W X(a，b) Substitute into the second judgment model to generate the label result Y t , and a second judgment result S2(t) is generated accordingly.

6. A coal blocking shutdown control system for a coal drop pipe according to claim 5, characterized in that: The shutdown control module comprises: A redundancy control unit connected to the result output unit, for receiving the first judgment result S1(t) and the second judgment result S2(t), and performing redundancy control by majority voting; that is, when S1(t)=S2(t), the output redundancy control result is S(t)=S1(t); when S1(t)≠S2(t), S2(t) is defined to take precedence over S1(t), and the output redundancy control result is S(t)=S2(t); An instruction generating unit, connected to the redundant control unit, and generating corresponding control instructions based on the redundant control result S(t); The shutdown control terminal is connected to the instruction generation unit, and is used to receive the control instruction and perform a shutdown operation corresponding to the content of the control instruction.

7. A coal blocking shutdown control system for a coal drop pipe according to claim 6, characterized in that: The alarm module comprises: A mobile device, which is wirelessly connected to the instruction generation unit and is used to display the control instruction corresponding to the current redundant control result in real time; The sound and light alarm unit is connected to the instruction generating unit and is arranged at the corresponding coal dropping pipe, and is used to perform sound and light alarm actions according to the control instructions.

8. A coal blocking shutdown control system for a coal drop pipe according to claim 7, characterized in that: During the training process, the support vector machine model in the machine learning unit adopts a cross-validation method to select the optimal parameter combination.

Images