Suspended climbing maintenance method for cyclone dust removal cavity of blast furnace

Abstract

The invention belongs to the technical field of blast furnace equipment maintenance, and provides a blast furnace cyclone dust removal cavity suspension climbing maintenance method which comprises the following steps: constructing a suspension type maintenance platform, constructing and deploying a three-dimensional perception network on the suspension type maintenance platform to collect multi-source monitoring data streams in real time, performing preliminary security state judgment in real time based on the multi-source monitoring data stream, if the preliminary security state is judged, triggering prospective multi-source monitoring data trend prediction, generating a predicted monitoring data sequence, and judging whether to trigger platform comprehensive evaluation, and if yes, performing prediction on the predicted multi-source monitoring data sequence; real-time multi-source monitoring data and prediction monitoring data sequences are fused on the basis of the D-S evidence theory, the comprehensive state of the platform is judged, and a progressive feedback mechanism or an ultimate feedback mechanism is started for the comprehensive state of the platform on the basis of the judgment result of the comprehensive state of the platform.

Description

Technical Field

[0001] This invention belongs to the field of blast furnace equipment maintenance technology, specifically a method for suspending and inspecting the blast furnace cyclone dust removal chamber at height. Background Technology

[0002] In blast furnace smelting systems, the cyclone dust collector chamber is a critical gas purification device. After long-term operation, it is prone to wear, corrosion, and ash accumulation, requiring regular high-altitude maintenance. However, the chamber's enclosed space and unique structure mean that the working platform is typically tens of meters above the bottom, creating a typical suspended and high-risk maintenance environment. Traditional maintenance methods often involve constructing temporary fixed scaffolding or welding supporting brackets to the chamber walls to build a working platform. These methods have revealed numerous technical limitations in practice.

[0003] First, the platform construction method has structural defects and equipment damage risks. Temporary support structures are not only time-consuming and labor-intensive to install, but may also fail under dynamic loads due to insufficient stress analysis or welding quality problems. On the other hand, safety status monitoring lacks foresight and comprehensiveness. Most existing technologies rely on manual periodic inspections or single sensor monitoring, which makes it difficult to capture the overall dynamic response of the platform system under complex operating loads, let alone predict stability trends. This leaves safety management in a passive mode of remediation after the fact, and the risk response mechanism is rigid and lagging. Existing methods lack the ability of graded early warning and proactive intervention. Emergency measures are usually only activated after a danger occurs. It is impossible to adjust the operation mode or apply stability control in advance according to subtle changes in the platform status, making it difficult to effectively curb the development of the accident chain.

[0004] Therefore, there is an urgent need to develop a suspended high-altitude maintenance method that can achieve non-destructive platform construction, integrate real-time multi-source monitoring, and have forward-looking risk prediction and intelligent hierarchical intervention capabilities, so as to fundamentally improve the safety, reliability and intelligence level of blast furnace cyclone dust removal chamber maintenance operations.

[0005] To address the above problems, this invention proposes a method for suspending and ascending the dust removal chamber of a blast furnace for maintenance. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve the technical problem is: a method for suspended and elevated maintenance of the blast furnace cyclone dust removal chamber, comprising:

[0008] Construct a suspended maintenance platform;

[0009] On a suspended maintenance platform, a three-dimensional sensing network is constructed and deployed to collect multi-source monitoring data streams in real time;

[0010] Based on the real-time multi-source monitoring data stream, a preliminary safety status is determined. If the preliminary safety status is determined, a forward-looking multi-source monitoring data trend prediction is triggered, a predicted monitoring data sequence is generated, and it is determined whether to trigger a comprehensive platform assessment.

[0011] If triggered, the system will fuse real-time multi-source monitoring data and predictive monitoring data sequences based on the DS evidence theory to determine the overall status of the platform.

[0012] Based on the overall status assessment results of the platform, a progressive feedback mechanism or a final feedback mechanism will be implemented to address the overall status of the platform.

[0013] Furthermore, the suspended maintenance platform is constructed as follows:

[0014] Outside the blast furnace cyclone dust collector chamber, multiple opening points are evenly determined and made according to the height position of the internal maintenance required. Multiple rigid support beams are inserted from the outside to the inside through the openings, leaving an outward extension outside the chamber and an inward extension inside the chamber. Inside the chamber, horizontal and vertical support components are installed on the inward extension of the rigid support beams to form a stable portal frame support. The top of the vertical support components is fixedly connected to the top inner wall of the blast furnace cyclone dust collector chamber. Outside the chamber, the outward extension of the rigid support beams is anchored to the fixed structure outside the chamber. Platform plates are laid on the portal frame support to form a ring maintenance platform. Safety nets are suspended on the outer edge of the maintenance platform. Multiple ring maintenance platforms are erected at vertical intervals according to maintenance needs.

[0015] Furthermore, the preliminary safety status determination method is as follows:

[0016] Independent safety thresholds are preset for each parameter of the multi-source monitoring data. Each parameter includes strain value, tilt angle, vibration amplitude and total load. The multi-source monitoring data at the current moment is quickly compared with the corresponding safety threshold. If all parameters are within the corresponding independent safety threshold range, the suspended maintenance platform is determined to be in a preliminary safe state.

[0017] Furthermore, the predictive monitoring data sequence is generated as follows:

[0018] A short-term historical period is set up. For any parameter contained in the multi-source monitoring data, the time series sequence of the parameter within the short-term historical period is extracted and marked as the historical parameter sequence. Based on the historical parameter sequence, a linear regression model is used as the core prediction algorithm. The linear relationship between the historical parameter sequence and the time variable is fitted by the least squares method. The time series numerical sequence of each parameter in the short future period with the current time as the starting point and the duration unchanged is predicted. The results are then integrated to obtain the predicted monitoring data sequence.

[0019] Furthermore, the method for determining whether a platform's comprehensive assessment has been triggered is as follows:

[0020] Compare the predicted monitoring data sequence with the corresponding safety threshold;

[0021] If all parameters are within the corresponding independent safety threshold range, it is determined that the suspended maintenance platform has no obvious risk of exceeding the limit in the short future period, triggering a comprehensive platform assessment.

[0022] Furthermore, the method for determining the overall status of the platform is as follows:

[0023] The decision-level fusion algorithm, based on Dempster evidence theory, treats real-time multi-source monitoring data and predictive monitoring data sequences as two independent evidence bodies. It defines an identification framework consisting of mutually exclusive and complete propositions, including safety, attention, and danger. Basic probability assignment functions are constructed for each of the two evidence bodies, and basic probability assignments are assigned. The Dempster synthesis rule is used to orthogonally fuse the two basic probability assignment functions, and the fused joint assignment is output as the proposition's confidence level. The proposition with the highest confidence level is selected as the determination result of the platform's overall status.

[0024] Furthermore, the allocation method for basic probability assignment is as follows:

[0025] For real-time multi-source monitoring data, based on the proximity of the values ​​of each parameter to the corresponding independent safety threshold, the degree of support for each proposition in the identification framework is calculated through a preset membership function to assign basic probability values. For predictive monitoring data sequences, based on the slope and stability of the changing trends of each parameter in the sequence, the tendency of future risk evolution is assessed and quantified as a basic probability assignment for each proposition in the identification framework.

[0026] Furthermore, the specific method of the dynamic progressive feedback mechanism is as follows:

[0027] Based on real-time multi-source monitoring data and predictive monitoring data sequences, a precise quantitative comprehensive evaluation value is calculated. Through the sound and light alarms deployed on each layer of the maintenance platform and the smart bracelets worn by the workers, based on the preset range of the comprehensive evaluation value, different levels of guidance information are generated and pushed, ranging from work behavior prompts to suggestions to suspend some high-risk operations.

[0028] Furthermore, the comprehensive evaluation value is calculated as follows:

[0029] A weighted fusion algorithm is adopted to dynamically weight and fuse the normalized values ​​of each parameter in the real-time multi-source monitoring data with the future maximum and minimum values ​​and the slope of the trend of each parameter in the predicted monitoring data sequence. A comprehensive evaluation value continuously distributed in the range of 0-100 is output through a preset evaluation function, where the larger the comprehensive evaluation value, the more significant the risk trend.

[0030] Furthermore, the specific method of the ultimate feedback mechanism is as follows:

[0031] The system activates the electric dampers deployed at key nodes of the portal frame support, which output a counterforce that precisely matches the platform's vibration modes to actively suppress abnormal shaking and deformation of the platform. It also immediately triggers the emergency deployment procedure of the winch-type safety net to quickly construct a continuous physical protective barrier below and to the side of the work area.

[0032] The beneficial effects of this invention are as follows:

[0033] 1. This invention integrates a three-dimensional sensing network with a DS evidence theory fusion algorithm to achieve a multi-dimensional, forward-looking comprehensive assessment of platform stability. This completely changes the traditional, lagging model that relies on manual experience and passive protection. Its core advantage is upgrading safety monitoring from static threshold alarms to dynamic trend warnings, enabling precise detection of early, weak signals of potential risks. This intelligent system not only significantly improves the depth and reliability of monitoring, effectively avoiding decision-making errors caused by false alarms from single sensors or incomplete data, but also significantly enhances the system's ability to cope with complex uncertainties through information fusion, setting a new safety standard for high-altitude, high-risk working environments.

[0034] 2. This invention introduces a hierarchical and progressive intelligent feedback mechanism, achieving precise and minimal safety intervention. By distinguishing between gradual and ultimate feedback, it can implement differentiated responses based on risk levels. When a state of alert is determined, the system provides guidance to workers through audible and visual alarms and smart bracelets, prompting them to adjust their work behavior autonomously. This non-coercive approach mitigates risks at their inception, ensuring the continuity of maintenance work. When a state of danger is determined, the system automatically triggers dampers and safety nets for mandatory physical intervention, constructing the ultimate safety defense line. This seamless mechanism, combining human and technical defenses and from early warning to physical protection, respects human initiative while ensuring reliable automated protection in emergencies, achieving the best balance between safety and efficiency. Attached Figure Description

[0035] The invention will now be further described with reference to the accompanying drawings.

[0036] Figure 1 This is a flowchart illustrating the steps of a method for suspending and inspecting a blast furnace cyclone dust removal chamber at height, as described in an embodiment of the present invention.

[0037] Figure 2 This is a flowchart illustrating the specific steps involved in determining the overall status of a blast furnace cyclone dust removal chamber during elevated maintenance, as described in an embodiment of the present invention. Detailed Implementation

[0038] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0039] Example 1

[0040] Please see Figure 1 As shown in the embodiment of the present invention, a method for suspended and elevated maintenance of a blast furnace cyclone dust collector includes the following steps:

[0041] S1: Construct a suspended maintenance platform;

[0042] Specifically, on the outside of the blast furnace cyclone dust removal chamber, multiple opening points are evenly determined along the circumference of the chamber shell according to the height position that needs to be inspected inside. Opening equipment is used to make holes at the marked points, and multiple rigid support beams are inserted from the outside to the inside through the openings. The rigid support beams retain an outward extension on the outside of the chamber and form an inward extension inside the chamber. Inside the chamber, horizontal support components and vertical support components are installed on the inward extension of the rigid support beams to form a stable portal-shaped support frame. The top of the vertical support components is fixedly connected to the top inner wall of the blast furnace cyclone dust removal chamber. On the outside of the chamber, the outward extension of the rigid support beams is anchored to the fixed structure outside the chamber. Platform plates are laid on the portal-shaped support frame to form a ring-shaped maintenance platform, and safety nets are suspended on the outer edge of the maintenance platform. According to maintenance needs, multiple ring-shaped maintenance platforms are erected at vertical intervals to complete the construction of the suspended maintenance platform.

[0043] S2: Construct and deploy a three-dimensional sensing network on a suspended maintenance platform to collect multi-source monitoring data streams in real time;

[0044] Specifically, based on the monitoring requirements for structural safety and stability during the suspended maintenance of the blast furnace cyclone dust removal chamber, a three-dimensional sensing network, including strain gauges, tilt sensors, vibration sensors, and weight sensors, is integrated and deployed on the key stress nodes of the portal-shaped support frame and the platform plate structure of the suspended maintenance platform.

[0045] Among them, the strain gauges collect the micro-bending change signals of the rigid support beam in real time at a preset sampling frequency based on the resistance strain effect; the tilt sensor senses the tilt angle change of the platform as a whole in the pitch and roll directions based on the built-in MEMS accelerometer component; the vibration sensor captures the platform structure vibration signal caused by the operation activities based on the piezoelectric effect; the weight sensor counts the real-time load borne by the platform based on the pressure sensing unit at the platform support point; and the three-dimensional sensing network collects data based on a unified time source and integrates them to form a heterogeneous original multi-source monitoring data vector.

[0046] Collaborative preprocessing is performed on the collected raw multi-source monitoring data vectors;

[0047] Specifically, a moving average filter is used to perform preliminary real-time filtering on each component of the original multi-source monitoring data vector. The window size is set to match the sampling frequency to smooth high-frequency noise introduced by instantaneous electromagnetic interference or random reading fluctuations, while retaining effective signals that characterize the changes in the actual state of the platform. Timestamp alignment of each component is performed. Based on a unified clock source, precise time stamps are inserted into the components from different sensors. The non-strictly synchronized data streams are unified to the same time reference through an interpolation algorithm. The aligned data is encapsulated according to a preset format specification to obtain multi-source monitoring data at each moment, including strain value, tilt angle, vibration amplitude and total load, forming a unified and time-synchronized multi-source monitoring data stream.

[0048] S3: Based on the multi-source monitoring data stream, perform a preliminary safety status determination in real time. If the determination is that the preliminary safety status is achieved, trigger a forward-looking multi-source monitoring data trend prediction, generate a predicted monitoring data sequence, and determine whether to trigger a comprehensive platform assessment.

[0049] Specifically, based on the safety requirements of suspended maintenance operations of the blast furnace cyclone dust removal chamber, independent safety thresholds are preset for each parameter of the multi-source monitoring data. Each parameter includes strain value, tilt angle, vibration amplitude and total load. The multi-source monitoring data at the current moment is quickly compared with the corresponding safety threshold. If all parameters are within the corresponding independent safety threshold range, the suspended maintenance platform is determined to be in a preliminary safe state. Otherwise, the suspended maintenance platform is determined to be in danger of exceeding the limit, triggering the emergency deployment procedure of the winch-type safety net to quickly construct a physical protective barrier below the work area and send an alarm to the administrator to report the danger of exceeding the limit.

[0050] It should be noted that the independent safety threshold is a hard boundary pre-set based on the platform structure design specifications, material mechanical properties and safety procedures of the suspended maintenance platform. Its function is to build a basic and rapid safety defense line to identify obvious over-limit or abnormal conditions in a timely manner and ensure that the system has the ability to respond to significant risks in a timely manner.

[0051] If the initial safety status is determined, a forward-looking multi-source monitoring data trend prediction will be triggered.

[0052] Specifically, a short historical period with the current time as the endpoint and a fixed length is set. For any parameter contained in the multi-source monitoring data, the time sequence of the parameter in the multi-source monitoring data stream within the short historical period is extracted and marked as a historical parameter sequence.

[0053] Based on historical parameter sequences, a linear regression model is used as the core prediction algorithm to extrapolate and predict the trends of each parameter in the multi-source monitoring data in a short future period with the current time as the starting point and the duration unchanged. The linear relationship between the historical parameter sequence and the time variable is fitted by the least squares method, thereby predicting the time series numerical sequence of each parameter in the short future period, and integrating them to obtain the predicted monitoring data sequence.

[0054] It should be noted that the linear regression model can extract the dominant trend of parameter changes from historical data. The prediction results quantify the expected evolution path of the platform's key stability parameters in the future if the current working conditions continue to develop. Its role is to expand the evaluation dimension of the system from a single current state to a spatiotemporal scope that includes the evolution of potential future risks, providing core data basis for forward-looking risk assessment.

[0055] The predicted monitoring data sequence is compared with the corresponding safety threshold. If all parameters are within the corresponding independent safety threshold range, it is determined that the suspended maintenance platform has no obvious risk of exceeding the limit in the short future period, triggering a comprehensive platform assessment. Otherwise, an alarm is sent to the administrator to report the risk of exceeding the limit.

[0056] It should be noted that the purpose of this step is to build a dual analysis mechanism that combines real-time judgment and forward-looking prediction capabilities. By comparing thresholds, it enables rapid screening of the platform's current security status, ensuring the system's basic response speed. On this basis, by using trend prediction based on historical data, it generates a predictive monitoring data sequence for assessing future risk evolution, thus providing a data foundation that is both real-time and predictive for subsequent comprehensive assessments, significantly improving the risk warning capabilities of the security monitoring system.

[0057] S4: If triggered, the platform's overall status is determined by fusing real-time multi-source monitoring data and predictive monitoring data sequences based on the DS evidence theory.

[0058] like Figure 2 As shown, the specific steps for determining the overall status of the platform are as follows;

[0059] Specifically, the system acquires real-time multi-source monitoring data at the current moment and predicted monitoring data sequences for short-term future periods. It uses the DS evidence theory as the core decision-level fusion algorithm to treat the real-time multi-source monitoring data and the predicted monitoring data sequences as two independent evidence bodies from different information sources.

[0060] It should be noted that the DS evidence theory is an important method for dealing with uncertainty. Its role is to provide a rigorous mathematical framework for the fusion of evidence from different information sources that may contain uncertainty, thereby improving the robustness and reliability of the comprehensive evaluation conclusion.

[0061] A recognition framework is constructed based on the DS evidence theory, and basic probability values ​​are assigned.

[0062] Specifically, a comprehensive assessment identification framework is defined, which consists of mutually exclusive and complete propositions, including safety, attention, and danger. Basic probability assignment functions are constructed for real-time multi-source monitoring data and predictive monitoring data sequences. For real-time multi-source monitoring data, based on the closeness of the values ​​of each parameter in the multi-source monitoring data to the corresponding independent safety threshold, the basic probability assignment is assigned by calculating the degree of support of the multi-source monitoring data for each proposition through a preset membership function. For predictive monitoring data sequences, based on the slope and stability of the changing trends of each parameter in the predictive monitoring data sequence, the tendency of future risk evolution is assessed and quantified into basic probability assignments for each proposition.

[0063] Evidence fusion and decision-making are performed based on Dempster's synthesis rules to determine the overall status of the platform;

[0064] Specifically, two basic probability assignment functions from real-time multi-source monitoring data and predictive monitoring data sequences are used as inputs. Orthogonal fusion is performed using the Dempster synthesis rule. The Dempster synthesis rule calculates the sum of the product of the support of the two evidence bodies for the same proposition and normalizes it to resolve conflicts between the evidence. Finally, the fused joint assignment is output as the confidence level of the proposition. The proposition with the highest confidence level is selected as the judgment result of the platform's overall status.

[0065] It should be noted that the purpose of this step is to construct a multi-source information decision fusion mechanism based on rigorous mathematical theory. By adopting the DS evidence theory, the system deeply integrates the immediacy advantage of real-time monitoring with the trend advantage of prediction data at the decision level, effectively handling the inherent uncertainty in monitoring and prediction information, thereby outputting a more robust and reliable overall platform status.

[0066] S5: Based on the overall platform status assessment results, implement a progressive feedback mechanism or an ultimate feedback mechanism to address the overall platform status.

[0067] Specifically, when the platform's overall status is determined to be "attention", a progressive feedback mechanism is activated. Based on real-time multi-source monitoring data and predictive monitoring data sequences, a precise quantitative comprehensive evaluation value is calculated. A weighted fusion algorithm is used to dynamically weight and fuse the normalized values ​​of each parameter in the real-time multi-source monitoring data with the future maximum and minimum values ​​and the slope of the changing trend of each parameter in the predictive data sequence. A comprehensive evaluation value that is continuously distributed in the range of 0-100 is output through a preset evaluation function.

[0068] It should be noted that a higher overall assessment value indicates a more significant risk trend;

[0069] By deploying sound and light alarms on each layer of the maintenance platform and using heterogeneous human-machine interaction media such as smart bracelets worn by workers, based on the comprehensive assessment value and the preset range of the comprehensive assessment value, different levels of guidance information are generated and pushed, ranging from work behavior prompts to suggestions to suspend some high-risk operations. The aim is to prompt workers to adjust their work intensity, optimize load distribution, or temporarily stop specific processes, so as to proactively intervene and resolve potential risks from the source.

[0070] It should be noted that the role of progressive feedback is to build a flexible and guided primary security defense line. Its core lies in relying on and enhancing personnel's autonomous risk awareness and avoidance ability through precise information prompts and early warnings, so as to achieve non-mandatory early risk control and intervention.

[0071] When the overall status of the platform is determined to be dangerous, the ultimate feedback mechanism is activated to directly control the pre-integrated active stabilization and safety protection devices to perform physical intervention. This triggers the electric dampers deployed at key nodes of the portal frame, which output a reverse force that precisely matches the platform's vibration modes to actively suppress abnormal shaking and deformation of the platform. It also immediately triggers the emergency deployment procedure of the winch-type safety net to quickly construct a continuous physical protective barrier below and to the side of the work area to deal with potential instability risks.

[0072] It should be noted that ultimate feedback is a mandatory safety measure that the system autonomously decides and executes. Its role is to actively apply stabilizing force and quickly build physical defenses when the risk exceeds the scope of immediate human control or when early warning guidance fails, thereby forcibly blocking the further development of the risk chain and ensuring the ultimate bottom line of personnel and equipment safety.

[0073] It should be noted that this step completes the entire safety closed loop of three-dimensional perception, prediction and assessment, intelligent decision-making, and precise execution. By transforming the judgment results of the platform's comprehensive status into hierarchical and precise feedback control from early warning guidance to physical intervention, a risk response and execution system with human-machine collaboration and timely response is constructed. This not only improves the proactive protection capability and intelligence level of the safety monitoring system, but also ensures that the risks are always within a knowable, controllable, and acceptable range throughout the entire high-altitude maintenance operation. This is the ultimate core guarantee for realizing the safety benefits of this method.

[0074] The technical solution of this invention is as follows: a suspended maintenance platform is built, and a three-dimensional sensing network is constructed and deployed on the suspended maintenance platform to collect multi-source monitoring data streams in real time. Based on the multi-source monitoring data streams, a preliminary safety status determination is performed in real time. If the preliminary safety status is determined, a forward-looking multi-source monitoring data trend prediction is triggered to generate a predicted monitoring data sequence. It is then determined whether a comprehensive platform assessment is triggered. If triggered, the real-time multi-source monitoring data and the predicted monitoring data sequence are fused based on the DS evidence theory to determine the comprehensive status of the platform. Based on the comprehensive status determination result, a progressive feedback mechanism or a final feedback mechanism is implemented to initiate the comprehensive status of the platform.

[0075] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber, characterized in that: include: Construct a suspended maintenance platform; On a suspended maintenance platform, a three-dimensional sensing network is constructed and deployed to collect multi-source monitoring data streams in real time; Based on the real-time multi-source monitoring data stream, a preliminary safety status is determined. If the preliminary safety status is determined, a forward-looking multi-source monitoring data trend prediction is triggered, a predicted monitoring data sequence is generated, and it is determined whether to trigger a comprehensive platform assessment. If triggered, the system will fuse real-time multi-source monitoring data and predictive monitoring data sequences based on the DS evidence theory to determine the overall status of the platform. Based on the overall platform status assessment results, a progressive feedback mechanism or a final feedback mechanism will be implemented to address the overall platform status.

2. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 1, characterized in that: The suspended maintenance platform is constructed as follows: Outside the blast furnace cyclone dust collector chamber, multiple opening points are evenly determined and made according to the height position requiring internal maintenance along the circumference of the chamber shell. Multiple rigid support beams are inserted from the outside to the inside through the openings, leaving an outward extension outside the chamber and an inward extension inside the chamber. Inside the chamber, horizontal and vertical support components are installed on the inward extension of the rigid support beams to form a stable portal frame support. The top of the vertical support components is fixedly connected to the top inner wall of the blast furnace cyclone dust collector chamber. Outside the chamber, the outward extension of the rigid support beams is anchored to the fixed structure outside the chamber. Platform plates are laid on the portal frame support to form a ring-shaped maintenance platform. Safety nets are suspended on the outer edge of the maintenance platform. Multiple ring-shaped maintenance platforms are erected at vertical intervals according to maintenance needs.

3. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 1, characterized in that: The preliminary safety status determination method is as follows: Independent safety thresholds are preset for each parameter of the multi-source monitoring data. Each parameter includes strain value, tilt angle, vibration amplitude and total load. The multi-source monitoring data at the current moment is quickly compared with the corresponding safety threshold. If all parameters are within the corresponding independent safety threshold range, the suspended maintenance platform is determined to be in a preliminary safe state.

4. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 3, characterized in that: The predictive monitoring data sequence is generated as follows: A short-term historical period is set up. For any parameter contained in the multi-source monitoring data, the time series sequence of the parameter within the short-term historical period is extracted and marked as the historical parameter sequence. Based on the historical parameter sequence, a linear regression model is used as the core prediction algorithm. The linear relationship between the historical parameter sequence and the time variable is fitted by the least squares method to predict the time series numerical sequence of each parameter in a short future period with the current time as the starting point and the duration unchanged. The results are then integrated to obtain the predicted monitoring data sequence.

5. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 4, characterized in that: The method for determining whether a platform's comprehensive assessment has been triggered is as follows: Compare the predicted monitoring data sequence with the corresponding safety threshold; If all parameters are within their respective independent safety threshold ranges, it is determined that the suspended maintenance platform has no significant risk of exceeding limits in the short future period, triggering a comprehensive platform assessment.

6. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 1, characterized in that: The method for determining the overall status of the platform is as follows: The decision-level fusion algorithm, based on Dempster evidence theory, treats real-time multi-source monitoring data and predictive monitoring data sequences as two independent evidence bodies. It defines an identification framework consisting of mutually exclusive and complete propositions, including safety, attention, and danger. Basic probability assignment functions are constructed for each of the two evidence bodies, and basic probability assignments are assigned. The Dempster synthesis rule is used to orthogonally fuse the two basic probability assignment functions, and the fused joint assignment is output as the proposition's confidence level. The proposition with the highest confidence level is selected as the determination result of the platform's overall status.

7. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 6, characterized in that: The basic probability assignment method is as follows: For real-time multi-source monitoring data, based on the proximity of the values ​​of each parameter to the corresponding independent safety threshold, the degree of support for each proposition in the identification framework is calculated through a preset membership function to assign basic probability values. For predictive monitoring data sequences, based on the slope and stability of the changing trends of each parameter in the sequence, the tendency of future risk evolution is assessed and quantified as a basic probability assignment for each proposition in the identification framework.

8. The method for suspended and elevated maintenance of the blast furnace cyclone dust collector chamber according to claim 1, characterized in that: The specific method of the dynamic progressive feedback mechanism is as follows: Based on real-time multi-source monitoring data and predictive monitoring data sequences, a precise quantitative comprehensive evaluation value is calculated. Through audible and visual alarms deployed on each layer of the maintenance platform and smart bracelets worn by operators, different levels of guidance information, ranging from work behavior prompts to suggestions to suspend some high-risk operations, are generated and pushed based on the preset range of the comprehensive evaluation value.

9. A method for suspended and elevated maintenance of a blast furnace cyclone dust collector chamber according to claim 8, characterized in that: The comprehensive evaluation value is calculated as follows: A weighted fusion algorithm is adopted to dynamically weight and fuse the normalized values ​​of each parameter in the real-time multi-source monitoring data with the future maximum and minimum values ​​and the slope of the trend of each parameter in the predicted monitoring data sequence. A comprehensive evaluation value continuously distributed in the range of 0-100 is output through a preset evaluation function, where the larger the comprehensive evaluation value, the more significant the risk trend.

10. A method for suspended and elevated maintenance of a blast furnace cyclone dust collector chamber according to claim 1, characterized in that: The specific method of the ultimate feedback mechanism is as follows: The system activates the electric dampers deployed at key nodes of the portal frame support, which output a counterforce that precisely matches the platform's vibration modes to actively suppress abnormal shaking and deformation of the platform. It also immediately triggers the emergency deployment procedure of the winch-type safety net to quickly construct a continuous physical protective barrier below and to the side of the work area.

Images