A coal blocking alarm method and system based on image recognition technology

Abstract

The application relates to a coal blocking alarm method and system based on image recognition technology in the technical field of coal conveying in a thermal power plant, which comprises the following steps: establishing a plurality of monitoring points according to equipment parameters in a plant area and setting an initial monitoring mode of the monitoring points; generating an image acquisition time node according to the initial monitoring mode, and acquiring image data of the monitoring points according to the image acquisition time node; establishing a coal level analysis model, generating real-time coal levels according to the image data of the monitoring points and the coal level analysis model; setting a monitoring mode of the monitoring points according to the real-time coal levels, generating a coal blocking early warning evaluation value, and judging whether to generate a maintenance instruction according to the coal blocking early warning evaluation value. The belt running image is shot in real time at the head position of each section of the coal conveying belt, the coal flow image at the head of the belt is analyzed in real time, and an alarm is sent when the coal level at the head of the belt exceeds the normal range, so that the coal blocking condition can be controlled in the initial stage.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal conveying in thermal power plants, in particular to a coal blocking alarm method and system based on image recognition technology. BACKGROUND

[0002] Coal blocking often occurs in the coal conveying system of a thermal power plant due to wet and sticky coal, which not only affects the normal operation of the coal conveying system, but also may cause the belt to deviate seriously, tear or break, and even damage the equipment. In daily operation, ordinary mechanical coal blocking switches are often not sensitive to coal blocking due to different working conditions and different severity of coal blocking, and cannot timely detect coal blocking. The belt inspector needs to inspect the equipment along the line, and cannot always monitor at the belt head. The coal conveying process control personnel cannot timely detect coal blocking due to the need to monitor multiple screens at the same time. A large number of manpower is needed to clean the coal blocking and restore the site after coal blocking occurs. SUMMARY

[0003] The purpose of the present application is to solve the above technical problems, and the present application provides a coal blocking alarm method and system based on image recognition technology, which aims to timely warn the coal blocking risk of the coal conveying system.

[0004] In some embodiments of the present application, the belt running image is captured in real time at the head position of each section of the coal conveying belt, the coal flow image at the belt head is analyzed in real time, and an alarm is issued when the coal accumulation at the belt head exceeds the normal range, thereby avoiding the disadvantages of the mechanical coal blocking switch being not sensitive and the limited manual effort being unable to accurately and timely detect the abnormality of the monitoring screen, and the coal blocking can be controlled in the initial stage.

[0005] In some embodiments of the present application, a coal blocking alarm method based on image recognition technology is provided, which comprises:

[0006] A plurality of monitoring points are established according to the equipment parameters of the plant, and an initial monitoring mode of the monitoring points is set;

[0007] An image acquisition time node is generated according to the initial monitoring mode, and image data of the monitoring points is acquired according to the image acquisition time node;

[0008] A coal level analysis model is established, and real-time coal level is generated according to the image data of the monitoring points and the coal level analysis model;

[0009] The monitoring mode of the monitoring points is set according to the real-time coal level, a coal blocking early warning evaluation value is generated, and a maintenance instruction is generated according to the coal blocking early warning evaluation value.

[0010] In some embodiments of the present application, when the image acquisition time node is generated according to the initial monitoring mode, it comprises:

[0011] When the initial monitoring mode is the first-level monitoring mode;

[0012] Obtaining historical operation data of the monitoring point, and obtaining a historical coal blocking frequency and an average coal blocking duration according to the historical operation data;

[0013] Generating a first historical evaluation value A1 according to the historical coal blocking frequency;

[0014] Generating a second historical evaluation value A2 according to the average coal blocking duration;

[0015] Generating a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2;

[0016] a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1;

[0017] Setting a time interval between adjacent image acquisition time nodes according to the historical evaluation value a.

[0018] In some embodiments of the present application, when establishing the coal level analysis model, the following steps are included:

[0019] Generating an image feature parameter according to historical image data;

[0020] Establishing an image feature parameter-coal level mapping table, and generating a training set data package and a verification set data package according to the image feature parameter-coal level mapping table;

[0021] Generating an initial coal level analysis model according to the training set data package through iterative training;

[0022] Generating a credibility of the initial coal level analysis model according to the verification set data package;

[0023] Setting a preset credibility threshold, and stopping iteration and outputting the coal level analysis model when the real-time credibility is greater than the credibility threshold.

[0024] In some embodiments of the present application, when setting the monitoring mode of the monitoring point according to the real-time coal level, the following steps are included:

[0025] Setting a first coal level threshold B1 and a second coal level threshold B2, and B1 < B2;

[0026] Obtaining a real-time coal level b;

[0027] If b < B1, setting the monitoring mode of the monitoring point as the first-level monitoring mode;

[0028] If B1 < b < B2, setting the monitoring mode of the monitoring point as the second-level monitoring mode, generating a coal blocking early warning evaluation value according to the second-level monitoring mode, and judging whether to generate a maintenance instruction according to the coal blocking early warning evaluation value.

[0029] If b > B2, the monitoring mode of the monitoring point is set to a third-level monitoring mode, and a maintenance instruction is generated.

[0030] In some embodiments of the present application, when generating the coal blockage early warning evaluation value according to the second-level monitoring mode, the following steps are included:

[0031] A feedback time node is set;

[0032] A feedback time node coal position sequence B is established, B=(b1, b2…bn), wherein n is the number of feedback time nodes, and bi is the real-time coal position of the i-th feedback time node;

[0033] A coal position change trend is generated according to the feedback time node coal position sequence B;

[0034] If the coal position change trend is a first-level trend, no coal blockage early warning evaluation value is generated, and when bi < B1, the monitoring mode of the monitoring point is corrected;

[0035] If the coal position change trend is a second-level trend, a coal position change speed c is generated according to the feedback time node coal position sequence B;

[0036] A coal blockage early warning evaluation value e of each feedback time node is generated according to the coal position change speed c;

[0037] A coal blockage early warning evaluation value threshold e1 is preset, and when ei < e1, a maintenance instruction is generated, wherein ei is the coal blockage early warning evaluation value of the i-th feedback time node.

[0038] If the coal position change trend is a third-level trend, a maintenance instruction is generated.

[0039] In some embodiments of the present application, when generating the coal blockage early warning evaluation value e according to the coal position change speed c, the following steps are included:

[0040] ei=[t2-t1]*E;

[0041] t2=(B2-bi) / ci

[0042] Wherein bi is the real-time coal position of the i-th feedback time node, ci is the coal position change speed of the i-th feedback time node, ei is the coal blockage early warning evaluation value of the i-th feedback time node, and t1 is the maintenance preparation time length.

[0043] In some embodiments of the present application, a coal blockage alarm system based on image recognition technology is provided, which includes:

[0044] A central control unit is configured to establish a plurality of monitoring points according to plant equipment parameters and set an initial monitoring mode of the monitoring points, and the central control unit is further configured to generate an image acquisition time node according to the initial monitoring mode.

[0045] a monitoring unit comprising a plurality of monitoring modules, the monitoring modules being arranged at the monitoring points, the monitoring modules being configured to acquire image data of the monitoring points according to the image acquisition time nodes;

[0046] the central control unit comprises:

[0047] a first processing module configured to establish a coal level analysis model, and generate real-time coal level according to the image data of the monitoring points and the coal level analysis model;

[0048] a second processing module configured to set a monitoring mode of the monitoring points according to the real-time coal level;

[0049] a third processing module configured to generate a coal blockage early warning evaluation value according to real-time image data of the monitoring points;

[0050] a maintenance module configured to generate a maintenance instruction.

[0051] In some embodiments of the present application, the central control unit further comprises:

[0052] a fourth processing module configured to, when the initial monitoring mode is a first monitoring mode;

[0053] the fourth processing module is configured to acquire historical operation data of the monitoring points, and acquire a historical coal blockage frequency and an average coal blockage duration according to the historical operation data;

[0054] generate a first historical evaluation value A1 according to the historical coal blockage frequency;

[0055] generate a second historical evaluation value A2 according to the average coal blockage duration;

[0056] generate a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2;

[0057] a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1;

[0058] set a time interval between adjacent image acquisition time nodes according to the historical evaluation value a.

[0059] In some embodiments of the present application, the second processing module is further configured to:

[0060] preset a first coal level threshold B1 and a second coal level threshold B2, and B1 < B2;

[0061] acquire a real-time coal level b;

[0062] if b < B1, the second processing module sets the monitoring mode of the monitoring points to a first monitoring mode;

[0063] If B1<b<B2, the second processing module sets the monitoring mode of the monitoring point to a second-level monitoring mode, the third processing module generates a coal blockage early warning evaluation value according to the second-level monitoring mode, and the maintenance module determines whether to generate a maintenance instruction according to the coal blockage early warning evaluation value;

[0064] If b>B2, the second processing module sets the monitoring mode of the monitoring point to a third-level monitoring mode, and the maintenance module generates a maintenance instruction.

[0065] In some embodiments of the present application, when the third processing module generates a coal blockage early warning evaluation value according to the second-level monitoring mode, the third processing module comprises:

[0066] setting a feedback time node;

[0067] establishing a feedback time node coal position sequence B, B=(b1, b2…bn), wherein n is the number of feedback time nodes, and bi is the real-time coal position of the i-th feedback time node;

[0068] generating a coal position change trend according to the feedback time node coal position sequence B;

[0069] If the coal position change trend is a first-level trend, no coal blockage early warning evaluation value is generated, and when bi<B1, the monitoring mode of the monitoring point is corrected;

[0070] If the coal position change trend is a second-level trend, a coal position change speed c is generated according to the feedback time node coal position sequence B;

[0071] a coal blockage early warning evaluation value e of each feedback time node is generated according to the coal position change speed c;

[0072] ei=[t2-t1]*E;

[0073] t2=(B2-bi) / ci

[0074] wherein bi is the real-time coal position of the i-th feedback time node, ci is the coal position change speed of the i-th feedback time node, ei is the coal blockage early warning evaluation value of the i-th feedback time node, and t1 is the maintenance preparation time length;

[0075] a coal blockage early warning evaluation value threshold e1 is preset, and when ei

[0076] If the coal position change trend is a third-level trend, the maintenance module generates a maintenance instruction.

[0077] Compared with the prior art, the coal blockage alarm method and system based on image recognition technology have the beneficial effects that:

[0078] Real-time shooting of belt running image at head position of each section coal conveying belt, real-time analysis of coal flow image at belt head, alarm in case of coal accumulation at belt head exceeding normal range, avoiding mechanical coal blocking switch insensitivity and limited manual energy, unable to accurately and timely find abnormal monitoring screen, and coal blocking can be controlled in initial stage. BRIEF DESCRIPTION OF DRAWINGS

[0079] Figure 1 is a flow diagram of a coal blocking alarm method based on image recognition technology in a preferred embodiment of the present application. DETAILED DESCRIPTION

[0080] The specific embodiments of the present application will be further described in detail below in conjunction with the drawings and examples. The following examples are used to illustrate the present application, but not to limit the scope of the present application.

[0081] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0082] The terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0083] In the description of the present application, it should be noted that unless otherwise specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be fixed connection, or detachable connection, or integral connection; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through intermediate medium, or internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0084] As shown in Figure 1 The coal blocking alarm method based on image recognition technology in the preferred embodiment of the present application comprises:

[0085] S101: Establishing a plurality of monitoring points according to plant equipment parameters, and setting initial monitoring mode of monitoring points;

[0086] S102: generating an image acquisition time node according to the initial monitoring mode, and acquiring image data of the monitoring point according to the image acquisition time node;

[0087] S103: establishing a coal level analysis model, and generating a real-time coal level according to the image data of the monitoring point and the coal level analysis model;

[0088] S104: setting a monitoring mode of the monitoring point according to the real-time coal level, generating a coal blockage early warning evaluation value, and determining whether to generate a maintenance instruction according to the coal blockage early warning evaluation value.

[0089] Specifically, the head position of each section of the coal conveying belt is set as a monitoring point.

[0090] Specifically, when establishing the coal level analysis model, the following steps are included:

[0091] generating image feature parameters according to historical image data;

[0092] establishing an image feature parameter-coal level mapping table, and generating a training set data package and a verification set data package according to the image feature parameter-coal level mapping table;

[0093] iteratively training according to the training set data package to generate an initial coal level analysis model;

[0094] generating a credibility of the initial coal level analysis model according to the verification set data package;

[0095] a credibility threshold is preset, and when the real-time credibility is greater than the credibility threshold, the iteration is stopped and the coal level analysis model is output.

[0096] Specifically, based on image analysis technology, feature parameters in the image data of the head position of each section of the coal conveying belt are extracted, the coal level in the image is determined, thereby establishing an image feature parameter-coal level mapping table, generating a training set data, iteratively training, and finally generating a coal level analysis model.

[0097] Specifically, when generating the image acquisition time node according to the initial monitoring mode, the following steps are included:

[0098] when the initial monitoring mode is a first-level monitoring mode;

[0099] acquiring historical operation data of the monitoring point, and acquiring a historical coal blockage frequency and an average coal blockage duration according to the historical operation data;

[0100] generating a first historical evaluation value A1 according to the historical coal blockage frequency;

[0101] generating a second historical evaluation value A2 according to the average coal blockage duration;

[0102] generating a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2;

[0103] a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1;

[0104] The time interval between the adjacent image acquisition time nodes is set according to the historical operation evaluation value a.

[0105] Specifically, the first historical evaluation value and the second historical evaluation value have the same value range, the more the historical coal blocking times, the higher the corresponding first historical evaluation value, the longer the evaluation coal blocking duration, the higher the corresponding second historical evaluation value, and the higher the historical operation evaluation value, which indicates that the risk of coal blocking at the head of the current coal conveying belt is higher, and the time interval between the adjacent image acquisition time nodes is smaller.

[0106] In the preferred embodiment of the present application, when the monitoring mode of the monitoring point is set according to the real-time coal level, it includes:

[0107] The preset first coal level threshold B1 and the second coal level threshold B2 are set, and B1 < B2;

[0108] The real-time coal level b is obtained;

[0109] If b < B1, the monitoring mode of the monitoring point is set to a first-level monitoring mode;

[0110] If B1 < b < B2, the monitoring mode of the monitoring point is set to a second-level monitoring mode, a coal blocking early warning evaluation value is generated according to the second-level monitoring mode, and whether to generate a maintenance instruction is determined according to the coal blocking early warning evaluation value;

[0111] If b > B2, the monitoring mode of the monitoring point is set to a third-level monitoring mode, and a maintenance instruction is generated.

[0112] Specifically, the first coal level threshold and the second coal level threshold can be set according to historical operation data, the first coal level threshold refers to a safe operation coal level, and the allowable coal level fluctuation between the first coal level threshold and the second coal level threshold is the second coal level threshold. When the second coal level threshold is exceeded, it indicates that there is a risk of coal blocking.

[0113] Specifically, the monitoring device is arranged at the head of each section of the coal conveying belt to capture the belt operation image in real time, when in the first-level monitoring mode, although the head of each section of the coal conveying belt is continuously acquired, the single image data of the corresponding node is obtained according to the image acquisition time node for analysis, thereby saving the processing load of the central control unit and improving the data processing efficiency.

[0114] Specifically, when in the secondary monitoring mode, it indicates that there is a risk of coal blockage at the head of the coal conveying belt, at this time, the head of the coal conveying belt is continuously collected and continuously analyzed according to the preset feedback time node, and the coal blockage risk is timely warned to ensure the operation efficiency of the coal conveying system.

[0115] Specifically, the time interval between adjacent image collection time nodes is much larger than the time interval between adjacent feedback time nodes.

[0116] Specifically, when generating the coal blockage early warning evaluation value according to the secondary monitoring mode, the following steps are included:

[0117] Set the feedback time node;

[0118] Establish the feedback time node coal position sequence B, B=(b1, b2…bn), wherein n is the number of feedback time nodes, and bi is the real-time coal position of the i-th feedback time node;

[0119] Generate the coal position change trend according to the feedback time node coal position sequence B;

[0120] If the coal position change trend is a primary trend, do not generate the coal blockage early warning evaluation value, and when bi

[0121] If the coal position change trend is a secondary trend, generate the coal position change speed c according to the feedback time node coal position sequence B;

[0122] Generate the coal blockage early warning evaluation value e of each feedback time node according to the coal position change speed c;

[0123] Pre-set the coal blockage early warning evaluation value threshold e1, and when ei

[0124] If the coal position change trend is a tertiary trend, generate the maintenance instruction.

[0125] Specifically, the primary change trend is a downward trend, and when the real-time coal position falls below the first coal position threshold, the monitoring mode of the monitoring point is corrected to the primary monitoring mode.

[0126] Specifically, the secondary change trend is an upward trend, and the tertiary change trend is a fluctuation trend, when in the tertiary change trend, there may be an operation risk, and maintenance should be performed in time.

[0127] Specifically, when generating the coal blockage early warning evaluation value e according to the coal position change speed c, the following steps are included:

[0128] ei=[t2-t1]*E;

[0129] t2=(B2-bi) / ci

[0130] bi, ci, ei, t1

[0131] According to the coal blockage warning method based on image recognition technology in any one of the preferred embodiments, the coal blockage warning system based on image recognition technology comprises:

[0132] The central control unit is configured to establish a plurality of monitoring points according to the device parameters of the plant area and set an initial monitoring mode for the monitoring points, and is further configured to generate image acquisition time nodes according to the initial monitoring mode.

[0133] The monitoring unit comprises a plurality of monitoring modules, and the monitoring modules are arranged at the monitoring points. The monitoring modules are configured to acquire image data of the monitoring points according to the image acquisition time nodes.

[0134] Specifically, the monitoring points are the head positions of the coal conveying belts, and the monitoring modules are preferably monitoring devices, so as to capture the running images of the belts in real time.

[0135] The central control unit comprises:

[0136] The first processing module is configured to establish a coal level analysis model and generate real-time coal levels according to the image data of the monitoring points and the coal level analysis model.

[0137] The second processing module is configured to set the monitoring modes of the monitoring points according to the real-time coal levels.

[0138] The third processing module is configured to generate coal blockage early warning evaluation values according to the real-time image data of the monitoring points.

[0139] The maintenance module is configured to generate maintenance instructions.

[0140] The fourth processing module is configured to acquire historical running data of the monitoring points, acquire the historical coal blockage times and the average coal blockage duration according to the historical running data, generate a first historical evaluation value A1 according to the historical coal blockage times, generate a second historical evaluation value A2 according to the average coal blockage duration, generate a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2, and set the initial monitoring mode to the first monitoring mode when the historical evaluation value a is less than a preset threshold value.

[0141] The fourth processing module is configured to acquire historical running data of the monitoring points, acquire the historical coal blockage times and the average coal blockage duration according to the historical running data, generate a first historical evaluation value A1 according to the historical coal blockage times, generate a second historical evaluation value A2 according to the average coal blockage duration, generate a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2, and set the initial monitoring mode to the first monitoring mode when the historical evaluation value a is less than a preset threshold value.

[0142] The first historical evaluation value A1 is generated according to the historical coal blockage times.

[0143] The second historical evaluation value A2 is generated according to the average coal blockage duration.

[0144] The historical evaluation value a is generated according to the first historical evaluation value A1 and the second historical evaluation value A2.

[0145] a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1.

[0146] The time interval between adjacent image acquisition time nodes is set according to the historical operation evaluation value a.

[0147] Specifically, the second processing module is further configured to:

[0148] The first coal level threshold B1 and the second coal level threshold B2 are preset, and B1 < B2;

[0149] Obtain the real-time coal level b;

[0150] If b < B1, the second processing module sets the monitoring mode of the monitoring point to the first-level monitoring mode;

[0151] If B1 < b < B2, the second processing module sets the monitoring mode of the monitoring point to the second-level monitoring mode, the third processing module generates a coal blockage early warning evaluation value according to the second-level monitoring mode, and the maintenance module determines whether to generate a maintenance instruction according to the coal blockage early warning evaluation value;

[0152] If b > B2, the second processing module sets the monitoring mode of the monitoring point to the third-level monitoring mode, and the maintenance module generates a maintenance instruction.

[0153] Specifically, when the third processing module generates the coal blockage early warning evaluation value according to the second-level monitoring mode, the third processing module comprises:

[0154] Set a feedback time node;

[0155] Establish a feedback time node coal level sequence B, B = (b1, b2…bn), wherein n is the number of feedback time nodes, and bi is the real-time coal level of the i-th feedback time node;

[0156] Generate a coal level change trend according to the feedback time node coal level sequence B;

[0157] If the coal level change trend is a first-level trend, no coal blockage early warning evaluation value is generated, and the monitoring mode of the monitoring point is corrected when bi < B1;

[0158] If the coal level change trend is a second-level trend, generate a coal level change speed c according to the feedback time node coal level sequence B;

[0159] Generate a coal blockage early warning evaluation value e of each feedback time node according to the coal level change speed c;

[0160] ei = [t2-t1]*E;

[0161] t2 = (B2-bi) / ci

[0162] Wherein, bi is the real-time coal position of the i-th feedback time node, ci is the coal position change speed of the i-th feedback time node, ei is the coal blockage early warning evaluation value of the i-th feedback time node, t1 is the maintenance preparation time length;

[0163] The preset coal blockage early warning evaluation value threshold e1, when ei

[0164] If the coal position change trend is a three-level trend, the maintenance module generates a maintenance instruction.

[0165] In some embodiments of the present application, the belt running image is photographed in real time at the head position of each section of the coal conveying belt, the coal flow image at the belt head is analyzed in real time, and an alarm is sent when the coal position accumulation exceeds the normal range, avoiding the disadvantages that the mechanical coal blockage switch is not sensitive and the human energy is limited, and the monitoring screen cannot be accurately and timely found. The disadvantages of abnormality, the coal blockage situation can be controlled in the initial stage.

[0166] The above only describes the preferred embodiments of the present application, and it should be pointed out that for ordinary skilled persons in the technical field, several improvements and replacements can be made without departing from the technical principles of the present application, and these improvements and replacements should also be considered as the protection scope of the present application.

Claims

1. A coal blocking alarm method based on image recognition technology, characterized in that, The method comprises the following steps: establishing multiple monitoring points according to plant equipment parameters and setting initial monitoring modes of the monitoring points; generating image acquisition time nodes according to the initial monitoring modes and acquiring image data of the monitoring points according to the image acquisition time nodes; establishing a coal level analysis model and generating real-time coal levels according to the image data of the monitoring points and the coal level analysis model; setting monitoring modes of the monitoring points according to the real-time coal levels and generating a coal blockage early warning evaluation value, and determining whether to generate a maintenance instruction according to the coal blockage early warning evaluation value; when the image acquisition time nodes are generated according to the initial monitoring modes, the method comprises the following steps: when the initial monitoring mode is a first-level monitoring mode; acquiring historical operation data of the monitoring points, acquiring a historical coal blockage frequency and an average coal blockage duration according to the historical operation data; generating a first historical evaluation value A1 according to the historical coal blockage frequency; generating a second historical evaluation value A2 according to the average coal blockage duration; generating a historical evaluation value a according to the first historical evaluation value A1 and the second historical evaluation value A2; a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1; setting a time interval between adjacent image acquisition time nodes according to the historical operation evaluation value a; when the monitoring modes of the monitoring points are set according to the real-time coal levels, the method comprises the following steps: presetting a first coal level threshold B1 and a second coal level threshold B2, and B1 < B2; acquiring a real-time coal level b; if b < B1, setting the monitoring mode of the monitoring point to a first-level monitoring mode; if B1 < b < B2, setting the monitoring mode of the monitoring point to a second-level monitoring mode, generating a coal blockage early warning evaluation value according to the second-level monitoring mode, and determining whether to generate a maintenance instruction according to the coal blockage early warning evaluation value; if b > B2, setting the monitoring mode of the monitoring point to a third-level monitoring mode and generating a maintenance instruction; when the coal blockage early warning evaluation value is generated according to the second-level monitoring mode, the method comprises the following steps: setting a feedback time node; establishing a feedback time node coal level sequence B, B = (b1, b2…bn), wherein n is the number of feedback time nodes, and bi is a real-time coal level of the i-th feedback time node; generating a coal level change trend according to the feedback time node coal level sequence B; if the coal level change trend is a first-level trend, no coal blockage early warning evaluation value is generated, and the monitoring mode of the monitoring point is corrected when bi < B1; if the coal level change trend is a second-level trend, a coal level change speed c is generated according to the feedback time node coal level sequence B; a coal blockage early warning evaluation value e of each feedback time node is generated according to the coal level change speed c; a coal blockage early warning evaluation value threshold e1 is preset, and a maintenance instruction is generated when ei < e1, wherein ei is the coal blockage early warning evaluation value of the i-th feedback time node; if the coal level change trend is a third-level trend, a maintenance instruction is generated.

2. The image recognition technology-based coal blocking alarm method of claim 1, wherein, when the coal level analysis model is established, the method comprises the following steps: generating image feature parameters according to historical image data; establishing an image feature parameter-coal level mapping table, generating a training set data package and a verification set data package according to the image feature parameter-coal level mapping table; iteratively training according to the training set data package to generate an initial coal level analysis model; According to the verification set data packet, a credibility of the initial coal level analysis model is generated; A credibility threshold is preset, and when the real-time credibility is greater than the credibility threshold, iteration is stopped, and a coal level analysis model is output.

3. A coal blockage alarm system based on image recognition technology, adopting the coal blockage alarm method based on image recognition technology in any one of claims 1-2, characterized in that, Comprise: The central control unit is configured to establish a plurality of monitoring points according to plant equipment parameters and set an initial monitoring mode of the monitoring points, and the central control unit is further configured to generate an image acquisition time node according to the initial monitoring mode; The monitoring unit comprises a plurality of monitoring modules, and the monitoring modules are arranged at the monitoring points, and the monitoring modules acquire image data of the monitoring points according to the image acquisition time node; The central control unit comprises: A first processing module configured to establish a coal level analysis model, generate a real-time coal level according to image data of the monitoring points and the coal level analysis model; A second processing module configured to set a monitoring mode of the monitoring points according to the real-time coal level; A third processing module configured to generate a coal blockage early warning evaluation value according to real-time image data of the monitoring points; A maintenance module configured to generate a maintenance instruction.

4. The image recognition technology based coal plugging alarm system of claim 3, wherein, The central control unit further comprises: A fourth processing module configured to, when the initial monitoring mode is a first-level monitoring mode; The fourth processing module acquires historical operation data of the monitoring points, acquires a historical coal blockage frequency and an average coal blockage duration according to the historical operation data; A first historical evaluation value A1 is generated according to the historical coal blockage frequency; A second historical evaluation value A2 is generated according to the average coal blockage duration; A historical evaluation value a is generated according to the first historical evaluation value A1 and the second historical evaluation value A2; a = n1*A1 + n2*A2, wherein n1 is a preset first weight coefficient, n2 is a preset second weight coefficient, and n1 + n2 = 1; A time interval between adjacent image acquisition time nodes is set according to the historical operation evaluation value a.

5. The image recognition technology based coal plugging alarm system of claim 4, wherein, The second processing module is further configured to: A first coal level threshold B1 and a second coal level threshold B2 are preset, and B1 < B2; A real-time coal level b is acquired; If b < B1, the second processing module sets the monitoring mode of the monitoring points to a first-level monitoring mode; If B1 < b < B2, the second processing module sets the monitoring mode of the monitoring points to a second-level monitoring mode, the third processing module generates a coal blockage early warning evaluation value according to the second-level monitoring mode, and the maintenance module determines whether to generate a maintenance instruction according to the coal blockage early warning evaluation value; If b > B2, the second processing module sets the monitoring mode of the monitoring points to a third-level monitoring mode, and the maintenance module generates a maintenance instruction.

Images