Dust removal equipment AI regulation and control system based on secondary metal smelting flame situation recognition

Abstract

The invention discloses a dust removal equipment AI regulation and control system based on regenerated metal smelting flame situation recognition, relates to the technical field of dust removal equipment AI regulation and control, and solves the technical problem that in the prior art, the equipment running state cannot be regulated and controlled in real time according to the working environment during metal smelting. The method specifically comprises the steps that after visual recognition image collection is normal, a visual recognition image collected at a preset point position is marked as a flame recognition image, flame height information and flame concentration information are obtained according to the flame recognition image, whether flame feature analysis is normal or not is deduced according to information comparison, and after the flame feature analysis is normal, flame recognition is performed. Flame swing parameters and spark quantity parameters are collected, whether the smelting reaction is sufficient or not is analyzed and inferred according to the parameters, and when all analysis results are generated in the visual recognition analysis process, an AI regulation and control system regulates and controls dust removal equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of AI control of dust removal equipment, and in particular to an AI control system of dust removal equipment based on flame situation recognition of recycled metal smelting. Background Art

[0002] Recycled metal smelting is an important way of resource recycling with significant economic and environmental benefits. Scrap metal recycling comes from a wide range of sources, such as industrial waste, waste equipment, scrapped cars, and scrapped household appliances. During the metal smelting process, a large amount of smoke and exhaust gas will be generated, causing serious harm to the environment and human health. Therefore, it is very necessary to install appropriate dust removal equipment.

[0003] The patent with announcement number CN115857587A discloses an intelligent control system for the operating environment of industrial dust removal equipment, including a designated dust removal space determination module, a monitoring point layout module, an effective monitoring area demarcation module, a dust distribution parameter collection module, an effective monitoring area ambient temperature determination module, a dust removal efficiency limit index analysis module, a control database, a key monitoring point identification module and a key monitoring point ambient temperature control module. It is through the layout of monitoring points in the designated dust removal space and the demarcation of the effective monitoring area corresponding to the monitoring points, and the collection of dust distribution parameters and ambient temperature at the corresponding monitoring points, so as to analyze the dust removal efficiency limit index, and then identify the key monitoring points, and perform temperature control at the key monitoring points, so as to realize the targeted control of the ambient temperature of the dust removal space where the dust removal equipment is located, greatly improve the adaptability of the control results and the control needs, and have more practical value.

[0004] However, the above content cannot perform characteristic analysis of the smelting flame based on visual recognition technology during metal smelting, so that the dust removal requirements of the smelting reaction cannot be inferred based on the real-time flame characteristics, and the dust removal equipment cannot be reasonably adjusted, and the effect of energy saving and emission reduction cannot be achieved.

[0005] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0006] The purpose of the present invention is to solve the above-mentioned problems and to propose an AI control system for dust removal equipment based on flame situation recognition of recycled metal smelting.

[0007] The purpose of the present invention can be achieved by the following technical solutions: an AI control system for dust removal equipment based on flame situation recognition of recycled metal smelting, the AI ​​control system is connected to a flame situation recognition platform, the AI ​​control system is used as a control terminal to control the operation of the dust removal equipment, and cooperates with the flame situation recognition platform to perform flame analysis on the real-time smelting process;

[0008] The flame situation recognition platform is connected with: a visual acquisition detection unit, which is used to collect and detect visual recognition acquisition, uniformly mark the flame acquisition points in the recycled metal smelting area as preset points, and detect the visual recognition image acquisition of the preset points, collect light impact data and dust impact data, and infer whether the visual recognition image acquisition is normal based on data comparison;

[0009] The flame feature analysis unit is used to mark the visual recognition image collected at the preset point as a flame recognition image after the visual recognition image is collected normally, and obtain the flame height information and flame concentration information based on the flame recognition image, and infer whether the flame feature analysis is normal based on the information comparison, and after it is normal, collect the flame swing parameters and spark quantity parameters, and infer whether the smelting reaction is sufficient based on the parameter analysis, and when each analysis result is generated during the visual recognition analysis process, the AI ​​control system controls the dust removal equipment.

[0010] As a preferred embodiment of the present invention, the light impact data and the dust impact data are respectively the cumulative duration of the ambient light reflection trajectory in the collection area passing through the location of the preset point when the preset point is continuously collected during the recycled metal smelting process, and the peak span of the increase in the area of ​​the occluded area in the collected image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process.

[0011] As a preferred embodiment of the present invention, if the light impact data exceeds the cumulative time threshold, or the dust impact data exceeds the area increase span peak threshold, a visual acquisition detection abnormal signal is generated; if the light impact data does not exceed the cumulative time threshold, and the dust impact data does not exceed the area increase span peak threshold, a visual acquisition detection normal signal is generated.

[0012] As a preferred embodiment of the present invention, the flame height information and the flame concentration information are respectively the cumulative values ​​of the speed corresponding to the speed of shortening the height interval value when the flame height does not exceed the set height when the smelting process needs to be in the heating stage at the same time node corresponding to the smelting process flow and the speed of increasing the height interval value when the flame height exceeds the set height, and the numerical ratio of the interval time when the flame concentration transparency changes from low to high and the continuous time when the flame concentration is at high transparency when the smelting process needs to be in the cooling stage at the same time node corresponding to the smelting process flow.

[0013] As a preferred embodiment of the present invention, if the flame height information exceeds the speed cumulative value threshold, or the flame concentration information exceeds the duration value ratio threshold, a smelting influence signal is generated and sent to the flame situation recognition platform. After the flame situation recognition platform receives the smelting influence signal, it is forwarded to the AI ​​control system.

[0014] If the flame height information does not exceed the speed cumulative value threshold, and the flame concentration information does not exceed the duration value ratio threshold, a normal smelting signal is generated and sent to the flame situation identification platform. After receiving the normal smelting signal, the flame situation identification platform transfers it to the AI ​​control system.

[0015] As a preferred embodiment of the present invention, the AI ​​control system sets the power of the dust removal equipment according to the required dust removal amount of the real-time smelting area, and automatically adjusts the set power according to the fluctuation of the dust removal amount; in addition, when there is no impact on smelting in the smelting area, auxiliary supply control is performed according to the real-time process impact; after the AI ​​control system receives the normal smelting signal, the dust removal equipment in the current smelting area is operated according to the power matched with the original set process flow.

[0016] As a preferred embodiment of the present invention, the flame swing parameter and the spark quantity parameter are respectively the sum of the span value of the instantaneous reduction of the flame swing amplitude in the high carbon stage of the smelting process in the flame recognition image and the corresponding span value of the flame height drop in the low carbon stage, and the increased span value of the flame bifurcation spark number in the recycled metal smelting stage in the flame recognition image.

[0017] As a preferred embodiment of the present invention, if the flame swing parameter exceeds the numerical value and the threshold, and the spark number parameter exceeds the span value threshold, a sufficient operation signal is generated; if the flame swing parameter does not exceed the numerical value and the threshold, or the spark number parameter does not exceed the span value threshold, an insufficient operation signal is generated and sent to the flame situation identification platform.

[0018] As a preferred embodiment of the present invention, after the flame situation recognition platform receives the insufficient operation signal, it is forwarded to the AI ​​control system. The AI ​​control system analyzes the gas products when the recycled metal smelting process is not fully reacted, and adjusts the power of the dust removal equipment according to the amount of gas products. In the gas product generation stage, the exhaust volume of the current period is increased. At the same time, when the amount of gas products does not reach the set standard, the dust removal equipment reduces the power and the exhaust gas is recovered through the smelting equipment, thereby reducing the operating energy consumption of the dust removal equipment.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] 1. In the present invention, visual recognition acquisition is collected and tested to infer whether the execution efficiency of visual acquisition images during recycled metal smelting meets the actual image acquisition requirements, so as to avoid the abnormal visual recognition acquisition environment causing deviations in visual recognition image acquisition, thereby causing the deviation of flame situation recognition to increase. In addition, according to real-time visual recognition analysis, the current visual acquisition efficiency can be accurately monitored in real time, which is convenient for point replacement in abnormal situations, ensuring that the flame type can be accurately analyzed during recycled metal smelting, so as to reasonably adjust the dust removal equipment, maximize energy conservation and emission reduction, and reduce the operating wear of various process equipment during the smelting process.

[0021] 2. In the present invention, the flame characteristics of the recycled metal smelting process are analyzed, and the flame characteristics are analyzed in combination with the recycled metal smelting stage to infer whether the corresponding smelting process execution in the current smelting stage meets the actual needs, so as to avoid the flame characteristics inferred by the flame characteristics failing to meet the actual smelting requirements, resulting in reduced smelting efficiency. At the same time, the original operation plan of the dust removal equipment has not changed with the progress of smelting, resulting in low control accuracy of the dust removal equipment, and unable to minimize the energy consumption and emissions of the smelting process, and unable to save energy and reduce emissions based on the basis;

[0022] While ensuring that the smelting process is qualified through flame characteristics, the flame recognition image is further analyzed to ensure that the dust removal equipment can operate more in line with the current smelting process progress. Improving the energy consumption control efficiency of the dust removal equipment can also minimize smelting gas emissions.

[0023] 3. In the present invention, auxiliary monitoring and analysis is performed on the recycled metal smelting process, specifically, the valve operating status and the amount of flue gas generated during the smelting process are monitored and analyzed. The dust removal equipment is intelligently regulated through auxiliary monitoring and analysis to ensure that the dust removal efficiency in the smelting area is not affected. The dust removal energy consumption can also be reduced by adjusting the operation of the dust removal equipment itself.

[0024] 4. In the present invention, the performance of the dust removal equipment is tested in the recycled metal smelting process. Through the equipment performance test analysis, it is inferred whether the dust removal equipment meets the dust removal requirements during actual operation, so as to facilitate the dust removal control of the entire area, while also ensuring the feasibility of intelligent regulation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0026] Figure 1 This is a principle block diagram of Embodiment 1 of the present invention;

[0027] Figure 2 This is a principle block diagram of Embodiment 2 of the present invention;

[0028] Figure 3This is a principle block diagram of embodiment 3 of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0031] Example 1: In this example, the dust removal equipment is accurately controlled according to the real-time progress analysis during the smelting of recycled metal. Figure 1 As shown, the AI ​​control system of dust removal equipment based on flame situation recognition of recycled metal smelting is connected to the flame situation recognition platform. As the control end, the AI ​​control system is used to control the operation of the dust removal equipment. In cooperation with the flame situation recognition platform, the flame analysis of the real-time smelting process is performed. The real-time flame state is analyzed according to the visual recognition technology, and the dust removal equipment is controlled according to the smelting conditions.

[0032] The flame situation recognition platform is connected to the visual acquisition detection unit and the flame feature analysis unit;

[0033] After the execution of recycled metal smelting, the flame situation recognition platform generates a visual acquisition detection signal and sends the visual acquisition detection signal to the visual acquisition detection unit. After receiving the visual acquisition detection signal, the visual acquisition detection unit performs acquisition detection on the visual recognition acquisition to infer whether the execution efficiency of the visual acquisition image during recycled metal smelting meets the actual image acquisition requirements, so as to avoid the visual recognition image acquisition deviation caused by the abnormal visual recognition acquisition environment, which may increase the deviation of flame situation recognition. In addition, according to the real-time visual recognition analysis, the current visual acquisition efficiency can be accurately monitored in real time, which is convenient for point replacement in abnormal situations, and ensures that the flame type can be accurately analyzed during recycled metal smelting, so as to reasonably adjust the dust removal equipment, maximize energy conservation and emission reduction, and reduce the operating wear of various process equipment during the smelting process;

[0034] The flame collection points in the recycled metal smelting area are uniformly marked as preset points, and the visual recognition image collection of the preset points is detected to obtain the cumulative time when the ambient light reflection trajectory of the collection area passes through the preset point position when the preset point is continuously collected during the recycled metal smelting process, and at the same time, the peak value of the increase span of the blocked area in the collection image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process is obtained. The cumulative time when the ambient light reflection trajectory of the collection area passes through the preset point position when the preset point is continuously collected during the recycled metal smelting process and the peak value of the increase span of the blocked area in the collection image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process are marked as light impact data and dust impact data, respectively, and are compared with the cumulative time threshold and the area increase span peak threshold, respectively:

[0035] If the accumulated time of the ambient light reflection trajectory in the collection area passing through the location of the preset point during continuous collection in the recycled metal smelting process exceeds the accumulated time threshold, or the peak value of the area increase span of the occluded area in the collected image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process exceeds the peak value of the area increase span threshold, then it is inferred that the visual recognition image collection in the recycled metal smelting area is abnormal, and a visual collection detection abnormality signal is generated and sent to the flame situation recognition platform. After receiving the visual collection detection abnormality signal, the flame situation recognition platform re-collects the current visual collection period and uses the already collected image as a reference image instead of a decision image.

[0036] If the accumulated time of the ambient light reflection trajectory in the collection area passing through the location of the preset point during continuous collection in the recycled metal smelting process does not exceed the accumulated time threshold, and the peak value of the area increase span of the occluded area in the collection image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process does not exceed the area increase span peak threshold, then it is inferred that the visual recognition image collection in the recycled metal smelting area is normal, and a visual collection detection normal signal is generated and sent to the flame situation recognition platform;

[0037] At the same time, a flame characteristic analysis signal is generated and sent to the flame characteristic analysis unit. After receiving the flame characteristic analysis signal, the flame characteristic analysis unit performs flame characteristic analysis on the recycled metal smelting process, and performs characteristic analysis on the flame in combination with the recycled metal smelting stage, so as to infer whether the execution of the corresponding smelting process in the current smelting stage meets the actual demand, so as to avoid the flame characteristics inferred by the flame characteristics failing to meet the actual smelting requirements, resulting in reduced smelting efficiency. At the same time, the operation plan of the original dust removal equipment has not changed with the progress of smelting, resulting in low control accuracy of the dust removal equipment, and failing to minimize the energy consumption and emissions of the smelting process, and failing to save energy and reduce emissions based on the results;

[0038] The visual recognition image collected at the preset point is marked as a flame recognition image, and the flame recognition image and the smelting process flow at the same time node in the smelting process are synchronously analyzed to obtain the corresponding speed cumulative value of the height interval value shortening speed when the flame height does not exceed the set height when the smelting process flow needs to be heated up at the same time node and the height interval value increasing speed when the flame height exceeds the set height. At the same time, the corresponding time node corresponding to the smelting process flow needs to be cooled down when the flame concentration transparency changes from low to high and the duration of the flame concentration at high transparency The corresponding time value ratio, wherein the flame transparency can be based on the object behind the flame as a reference, and the image is collected from the front of the flame, and the visibility of the reference object behind the flame in the collected image is used as the transparency value;

[0039] And the speed cumulative value corresponding to the speed of shortening the height interval value when the flame height does not exceed the set height when the smelting process needs to heat up at the same time node corresponding to the stage, and the speed cumulative value corresponding to the speed of increasing the height interval value when the flame height exceeds the set height, and the time value ratio corresponding to the interval time when the flame concentration transparency changes from low to high when the smelting process needs to cool down at the same time node corresponding to the stage, and the duration of the flame concentration at high transparency are marked as flame height information and flame concentration information, and compared with the speed cumulative value threshold and the duration value ratio threshold respectively:

[0040] If the corresponding cumulative value of the speed of shortening the height interval value when the flame height does not exceed the set height and the speed of increasing the height interval value when the flame height exceeds the set height at the same time node corresponding to the heating stage of the smelting process at the same time node exceeds the speed cumulative value threshold, or the corresponding time value ratio of the interval time when the flame concentration transparency changes from low to high and the duration of the flame concentration at high transparency at the same time node corresponding to the cooling stage of the smelting process exceeds the time value ratio threshold, it is inferred that the flame characteristic analysis of the corresponding time node is abnormal, and a smelting influence signal is generated and sent to the flame situation identification platform. After receiving the smelting influence signal, the flame situation identification platform transfers it to the AI ​​control system. The AI ​​control system sets the power of the dust removal equipment according to the required dust removal amount of the real-time smelting area and automatically adjusts the set power according to the fluctuation of the dust removal amount; in addition, when there is no impact on smelting in the smelting area, supply auxiliary control is performed according to the real-time process impact, such as replenishing raw materials or increasing the oxygen content.

[0041] If the corresponding cumulative value of the speed of shortening the height interval value when the flame height does not exceed the set height and the speed of increasing the height interval value when the flame height exceeds the set height at the same time node corresponding to the heating stage of the smelting process flow at the same time node does not exceed the speed cumulative value threshold, and the corresponding time value ratio of the interval time when the flame concentration transparency changes from low to high and the continuous time when the flame concentration is at high transparency when the smelting process flow at the same time node corresponds to the cooling stage does not exceed the time value ratio threshold, it is inferred that the flame characteristic analysis of the corresponding time node is normal, and a normal smelting signal is generated and sent to the flame situation recognition platform. After receiving the normal smelting signal, the flame situation recognition platform transfers it to the AI ​​control system. After receiving the normal smelting signal, the AI ​​control system operates the dust removal equipment in the current smelting area according to the power matching the original set process flow;

[0042] When ensuring that the smelting process is qualified through flame characteristics, further analysis of the flame recognition image can make the dust removal equipment run more in line with the current smelting process progress. Improving the energy consumption control efficiency of the dust removal equipment can also minimize smelting gas emissions.

[0043] The sum of the span values ​​of the instantaneous reduction in the flame swing amplitude in the high-carbon stage of the smelting process flow in the flame recognition image and the corresponding span values ​​of the span value of the flame height reduction in the low-carbon stage is obtained, wherein the span values ​​of different units are only calculated as numerical values, and the influence of different units is not considered; at the same time, the increased span value of the number of flame forked sparks in the recycled metal smelting stage in the flame recognition image is obtained, and the sum of the span values ​​of the instantaneous reduction in the flame swing amplitude in the high-carbon stage of the smelting process flow in the flame recognition image and the corresponding span values ​​of the span value of the flame height reduction in the low-carbon stage, and the increased span value of the number of flame forked sparks in the recycled metal smelting stage in the flame recognition image are marked as flame swing parameters and spark quantity parameters, respectively, and compared with the numerical value and threshold and span value threshold respectively: wherein, in the technical scheme, morphological data analysis of flames and sparks is performed at different smelting stages, and the spark flame characteristics are used to infer whether the current smelting reaction is sufficient, so as to predict in real time whether excess gas is generated, and reasonably regulate the dust removal equipment;

[0044] If the sum of the span values ​​corresponding to the instantaneous decrease in the span value of the flame swing amplitude in the high-carbon stage of the smelting process in the flame recognition image and the span value corresponding to the decrease in the flame height in the low-carbon stage exceeds the value and threshold, and the increase in the span value of the number of flame bifurcation sparks in the recycled metal smelting stage in the flame recognition image exceeds the span value threshold, it is inferred that the reaction of the current flame recognition image corresponding to the smelting period is sufficient, and a sufficient operation signal is generated and sent to the flame situation recognition platform. After receiving the sufficient operation signal, the flame situation recognition platform forwards it to the AI ​​control system, and the AI ​​control system sets the set power of the current dust removal equipment constant;

[0045] If the sum of the span values ​​corresponding to the instantaneous reduction in the span value of the flame swing amplitude in the high-carbon stage of the smelting process in the flame recognition image and the span value corresponding to the reduction in the flame height in the low-carbon stage does not exceed the value and threshold, or the increase in the span value of the number of flame bifurcation sparks in the recycled metal smelting stage in the flame recognition image does not exceed the span value threshold, it is inferred that the reaction of the smelting period corresponding to the current flame recognition image is insufficient, and an insufficient operation signal is generated and sent to the flame situation recognition platform. After receiving the insufficient operation signal, the flame situation recognition platform transfers it to the AI ​​control system. The AI ​​control system analyzes the gas products when the recycled metal smelting process is insufficiently reacted, and adjusts the power of the dust removal equipment according to the amount of gas products, and increases the exhaust volume of the current period in the gas product generation stage. At the same time, when the amount of gas products does not reach the set standard, the dust removal equipment reduces the power and the exhaust gas is recovered through the smelting equipment to reduce the operating energy consumption of the dust removal equipment;

[0046] Example 2: In the above-mentioned Example 1, the flame situation of the smelting process is identified, and the dust removal equipment is controlled according to the visual recognition result to achieve energy saving and emission reduction. Please refer to Figure 2 As shown, this embodiment is based on the previous embodiment, and an auxiliary monitoring and analysis unit is added;

[0047] After the flame characteristic analysis unit completes operation, it generates an auxiliary monitoring and analysis signal and sends the auxiliary monitoring and analysis signal to the auxiliary monitoring and analysis unit. After receiving the auxiliary monitoring and analysis signal, the auxiliary monitoring and analysis unit performs auxiliary monitoring and analysis on the recycled metal smelting process, specifically monitoring and analyzing the valve operation status and the amount of flue gas generated during the smelting process. Through the auxiliary monitoring and analysis, the dust removal equipment is intelligently controlled to ensure that the dust removal efficiency of the smelting area is not affected, and the dust removal energy consumption can be reduced through the operation adjustment of the dust removal equipment itself;

[0048] The deviation of the non-equal supply time corresponding to the opening amount of the control valve when the reaction of the smelting equipment progresses during the recycled metal smelting process and the actual demand is obtained, and the continuous increase span value of the smoke diffusion area in the smelting area when the dust removal equipment is running at the set power during the recycled metal smelting process is obtained, and the deviation of the non-equal supply time corresponding to the opening amount of the control valve when the reaction of the smelting equipment progresses during the recycled metal smelting process and the actual demand, and the continuous increase span value of the smoke diffusion area in the smelting area when the dust removal equipment is running at the set power during the recycled metal smelting process are compared with the supply time deviation threshold and the area continuous increase span threshold respectively: wherein the control valve is represented as a raw material delivery pipeline valve or a supply pipeline valve;

[0049] If the deviation of the supply time of the non-equal quantity corresponding to the control valve opening amount and the actual required amount during the reaction progress of the smelting equipment in the recycled metal smelting process exceeds the supply time deviation threshold, or the continuous increase span value of the smoke diffusion area in the smelting area when the dust removal equipment is running at the set power in the recycled metal smelting process exceeds the continuous increase span threshold, it is inferred that the auxiliary monitoring and analysis in the recycled metal smelting process is abnormal, and an auxiliary monitoring abnormal signal is generated and sent to the flame situation identification platform. After receiving the signal, the flame situation identification platform forwards it to the AI ​​control system. The AI ​​control system controls the dust removal equipment, automatically changes the original set power, and adjusts the power according to the real-time smoke volume in the smelting area and the increase in the smelting reaction products;

[0050] If the deviation of the non-equal supply time between the control valve opening amount and the actual required amount during the reaction progress of the smelting equipment in the recycled metal smelting process does not exceed the supply time deviation threshold, and the continuous increase span value of the fume diffusion area in the smelting area when the dust removal equipment is running at the set power during the recycled metal smelting process does not exceed the continuous increase span threshold, it is inferred that the auxiliary monitoring analysis in the recycled metal smelting process is normal, and an auxiliary monitoring normal signal is generated and sent to the flame situation recognition platform;

[0051] Example 3: Please refer to Figure 3 As shown, this embodiment adds an equipment performance detection platform on the basis of the previous two embodiments, which is used to perform performance detection on the dust removal equipment, so as to avoid the abnormality of the dust removal equipment during the smelting stage, which leads to a decrease in the dust removal efficiency in the smelting area, so that the dust removal equipment generates unnecessary AI regulation;

[0052] When the dust removal equipment is running continuously, the AI ​​control system generates an equipment performance detection signal and sends the equipment performance detection signal to the equipment performance detection unit. After receiving the equipment performance detection signal, the equipment performance detection unit performs a dust removal equipment performance test on the recycled metal smelting process. Through the equipment performance detection analysis, it is inferred whether the dust removal equipment meets the dust removal requirements during actual operation, so as to facilitate the dust removal control of the entire area and ensure the executability of intelligent control.

[0053] In the absence of environmental influences, the duration of the synchronous decrease of the maximum dust removal and filtration volume span of the filter surface of the dust removal equipment and the gas circulation velocity of the filter surface of the dust removal equipment in the recycled metal smelting process is obtained, and at the same time, the deviation value of the actual gas emission speed in the smelting area and the gas emission demand speed under the current dust removal performance is obtained when the performance of the filter surface of the dust removal equipment does not fluctuate in the recycled metal smelting process, and the deviation value of the actual gas emission speed in the smelting area and the gas emission demand speed under the current dust removal performance when the performance of the filter surface of the dust removal equipment does not fluctuate in the recycled metal smelting process are compared with the synchronous decrease duration threshold and the speed deviation value threshold respectively: wherein, no environmental influence means excluding the influences such as inconsistent dust removal demand on the filter surface of the dust removal equipment and the presence of dust in the dust removal environment;

[0054] If the maximum dust removal and filtration volume span of the filter surface of the dust removal equipment in the recycled metal smelting process and the duration of the synchronous decline of the gas circulation speed of the filter surface of the dust removal equipment exceed the synchronous decline duration threshold, or when the performance of the filter surface of the dust removal equipment does not fluctuate during the recycled metal smelting process, the deviation value between the actual gas emission speed in the smelting area and the gas emission demand speed under the current dust removal performance exceeds the speed deviation value threshold, it is inferred that the equipment performance detection in the recycled metal smelting process is abnormal, and an equipment control signal is generated and sent to the AI ​​control system. After receiving the equipment control signal, the AI ​​control system cleans and maintains the hardware components in the dust removal equipment and adaptively adjusts the dust removal power of the dust removal equipment;

[0055] If the maximum dust removal and filtration volume span of the filter surface of the dust removal equipment during the recycled metal smelting process and the duration of the synchronous decline in the gas circulation speed of the filter surface of the dust removal equipment do not exceed the synchronous decline duration threshold, and the performance of the filter surface of the dust removal equipment does not fluctuate during the recycled metal smelting process, the deviation between the actual gas emission speed in the smelting area and the required gas emission speed under the current dust removal performance does not exceed the speed deviation value threshold, then it is inferred that the equipment performance detection during the recycled metal smelting process is normal, and a qualified performance signal is generated and sent to the AI ​​control system.

[0056] When the present invention is in use, the visual acquisition detection unit performs acquisition and detection on the visual recognition acquisition, uniformly marks the flame acquisition points in the recycled metal smelting area as preset points, and detects the visual recognition image acquisition of the preset points, collects light impact data and dust impact data, and infers whether the visual recognition image acquisition is normal based on data comparison; after the visual recognition image acquisition is normal, the flame feature analysis unit marks the visual recognition image collected at the preset point as a flame recognition image, and obtains flame height information and flame concentration information based on the flame recognition image, infers whether the flame feature analysis is normal based on information comparison, and after it is normal, collects flame swing parameters and spark quantity parameters, infers whether the smelting reaction is sufficient based on parameter analysis, and when each analysis result is generated during the visual recognition analysis process, the AI ​​control system controls the dust removal equipment.

[0057] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. An AI control system for dust removal equipment based on flame situation recognition for recycled metal smelting. The AI ​​control system is connected to a flame situation recognition platform, and is characterized in that: As the control end, the AI ​​control system is used to control the operation of the dust removal equipment and cooperate with the flame situation recognition platform to perform flame analysis on the real-time smelting process; The flame situation recognition platform is connected with: a visual acquisition detection unit, which is used to collect and detect visual recognition acquisition, uniformly mark the flame acquisition points in the recycled metal smelting area as preset points, and detect the visual recognition image acquisition of the preset points, collect light impact data and dust impact data, and infer whether the visual recognition image acquisition is normal based on data comparison; The flame feature analysis unit is used to mark the visual recognition image collected at the preset point as a flame recognition image after the visual recognition image is collected normally, and obtain the flame height information and flame concentration information based on the flame recognition image, and infer whether the flame feature analysis is normal based on the information comparison, and after it is normal, collect the flame swing parameters and spark quantity parameters, and infer whether the smelting reaction is sufficient based on the parameter analysis, and when each analysis result is generated during the visual recognition analysis process, the AI ​​control system controls the dust removal equipment.

2. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 1 is characterized in that: The light impact data and dust impact data are respectively the cumulative duration of the ambient light reflection trajectory in the collection area passing through the location of the preset point when the preset point is continuously collected during the recycled metal smelting process, and the peak span of the increase in the occluded area in the collected image corresponding to any preset point in the area during the dust increase stage in the smelting area during the recycled metal smelting process.

3. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 2 is characterized in that: If the light impact data exceeds the cumulative duration threshold, or the dust impact data exceeds the area increase span peak threshold, a visual acquisition detection abnormal signal is generated; if the light impact data does not exceed the cumulative duration threshold, and the dust impact data does not exceed the area increase span peak threshold, a visual acquisition detection normal signal is generated.

4. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 1 is characterized in that: The flame height information and flame concentration information are respectively the cumulative values ​​of the speed corresponding to the shortening rate of the height interval value when the flame height does not exceed the set height when the smelting process needs to be in the heating stage at the same time node corresponding to the stage where the flame height is in the heating stage, and the speed corresponding to the increasing rate of the height interval value when the flame height exceeds the set height, and the numerical ratio of the interval time when the flame concentration transparency changes from low to high and the continuous time when the flame concentration is in high transparency when the smelting process needs to be in the cooling stage at the same time node corresponding to the stage where the flame height is in the heating stage.

5. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 4 is characterized in that: If the flame height information exceeds the speed cumulative value threshold, or the flame concentration information exceeds the duration value ratio threshold, a smelting influence signal is generated and sent to the flame situation recognition platform. After receiving the smelting influence signal, the flame situation recognition platform forwards it to the AI ​​control system. If the flame height information does not exceed the speed cumulative value threshold, and the flame concentration information does not exceed the duration value ratio threshold, a normal smelting signal is generated and sent to the flame situation identification platform. After receiving the normal smelting signal, the flame situation identification platform transfers it to the AI ​​control system.

6. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 5 is characterized in that: The AI ​​control system sets the power of the dust removal equipment according to the dust removal requirements of the real-time smelting area, and automatically adjusts the set power according to the fluctuation of the dust removal amount. In addition, when there is no impact on smelting in the smelting area, auxiliary control is performed based on the real-time process impact. After receiving the normal smelting signal, the AI ​​control system operates the dust removal equipment in the current smelting area according to the power that matches the original set process flow.

7. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 1 is characterized in that: The flame swing parameter and spark quantity parameter are respectively the sum of the span value of the instantaneous reduction of the flame swing amplitude in the high carbon stage of the smelting process in the flame recognition image and the corresponding span value of the flame height drop in the low carbon stage, and the increased span value of the flame bifurcation spark number in the recycled metal smelting stage in the flame recognition image.

8. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 7 is characterized in that: If the flame swing parameter exceeds the numerical value and threshold, and the spark number parameter exceeds the span value threshold, a full operation signal is generated; if the flame swing parameter does not exceed the numerical value and threshold, or the spark number parameter does not exceed the span value threshold, a non-full operation signal is generated and sent to the flame situation identification platform.

9. The AI ​​control system for dust removal equipment based on flame situation recognition for recycled metal smelting according to claim 8 is characterized in that: After receiving the insufficient operation signal, the flame situation recognition platform transfers it to the AI ​​control system. The AI ​​control system analyzes the gas products when the recycled metal smelting process is not fully reacted, and adjusts the power of the dust removal equipment according to the amount of gas products. In the gas product generation stage, the exhaust volume of the current period is increased. At the same time, when the amount of gas products does not reach the set standard, the dust removal equipment reduces the power and the exhaust gas is recovered through the smelting equipment, reducing the operating energy consumption of the dust removal equipment.

Images