Intelligent analysis method and system for damage of boiler heating surface
By dividing the boiler heating surface into multiple zones and establishing reference zones, and combining multiple diagnostic actions and parameter control, the inaccuracy of judging boiler heating surface damage in the existing technology has been solved, and the damage type can be accurately located and identified under complex operating conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 新疆宇澄热力股份有限公司
- Filing Date
- 2026-04-16
- Publication Date
- 2026-07-31
AI Technical Summary
Current methods for judging damage to boiler heating surfaces rely on single-point over-limit, static threshold, or trend analysis under a single operating condition. This makes it difficult to conduct regional comparisons under complex operating conditions, resulting in inaccurate damage location, unclear type identification, and insufficient stability of analysis results.
The boiler heating surface is divided into multiple analysis zones, and symmetrical reference zones, co-operation reference zones, and co-loop reference zones are established. Through multiple diagnostic actions and parameter control within the short-term diagnostic window, the net recovery amount, abnormal center migration amount, and relative hysteresis amount are determined. Combined with the retention and deletion of preset damage types, the accurate location of damage types is achieved.
It effectively distinguishes different sources of anomalies, solves the problems of unclear damage type identification and insufficient accuracy of analysis results under complex operating conditions, and realizes accurate location and type identification of boiler heating surface damage.
Smart Images
Figure CN122020498B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of boiler operation monitoring and fault diagnosis technology, specifically to an intelligent analysis method and system for boiler heating surface damage. Background Technology
[0002] Coal-fired boilers, as core thermal energy equipment in the steam supply systems of power plants, utility stations, and combined oil refining and oilfield stations, have long been responsible for steam generation, heating, and heat conversion. With the increasing size of thermal power units, the continuous operation of refining and chemical plants, and the intensive development of supporting thermal systems for oil and gas extraction, boiler operation has gradually evolved from relying on manual inspections and single-point instrument judgments to an operation mode based on multi-parameter acquisition, online monitoring, and automatic control. Real-time identification and analysis of the safety status of heating surfaces has also become an important requirement for ensuring the long-term stable operation of boilers.
[0003] Existing methods for judging boiler heating surface damage mostly rely on single-point over-limit, static threshold alarms, or trend analysis under single operating conditions. When faced with coupled operating conditions such as fluctuations in coal quality, soot blowing disturbances, load changes, and changes in steam demand in petroleum refining units, it is easy to confuse flue gas adhesion, combustion deviation, local flow abnormalities, insufficient cooling, and abnormalities at measuring points. Furthermore, it is difficult to compare and distinguish different analysis areas under a unified baseline, resulting in inaccurate damage location, unclear type identification, and insufficient stability of analysis results. Therefore, intelligent analysis methods and systems for boiler heating surface damage are needed to solve the above problems. Summary of the Invention
[0004] (a) Technical problems to be solved To address the shortcomings of existing technologies, this invention provides an intelligent analysis method and system for boiler heating surface damage, thus solving the aforementioned problems.
[0005] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: an intelligent analysis method for boiler heating surface damage, comprising: S1. Divide the heating surface of the target boiler into multiple analysis zones, and establish a symmetrical reference zone, a reference zone with the same action, and a reference zone with the same loop for each analysis zone; S2. Obtain the status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. S3. Within the short-term diagnostic window, control the key operating parameters of the boiler to be within the preset fluctuation range, and make each diagnostic action start from the same target baseline range; perform two short-term soot blowing actions with alternating order on the target analysis area and its corresponding reference area to determine the net recovery amount of the target analysis area; S4. After returning to the target baseline range, perform two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled to the target analysis area to determine the abnormal center migration amount; after returning to the target baseline range, perform small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis amount of the target analysis area relative to the same loop reference area. S5. Based on the net recovery amount, the abnormal center migration amount, and the relative backlash amount, retain or delete preset damage types, determine the target damage type, and output the corresponding analysis results.
[0006] Furthermore, S1 specifically includes: The target boiler's heating surface is divided into multiple analysis zones according to the medium flow path; For each of the analysis regions, an analysis region that is symmetrical to its position in the boiler structure is determined as the symmetry reference region; The analysis area subjected to the same soot blowing device is identified as the reference area for the same action; The analysis area located in the same header and the same parallel branch is determined as the reference area of the same loop.
[0007] Furthermore, S2 specifically includes: Obtain wall temperature data, flue gas temperature data, steam temperature data, flow rate data, and pressure data corresponding to each of the analysis zones; Based on the wall temperature data corresponding to each of the analysis zones and the wall temperature data corresponding to the symmetrical reference zone, the wall temperature deviation is determined. Based on the smoke temperature data corresponding to each analysis zone and the smoke temperature data corresponding to the same action reference zone, the smoke temperature deviation is determined; Based on the steam temperature, flow rate, and pressure data corresponding to each analysis zone and the steam temperature, flow rate, and pressure data corresponding to the reference zone in the same loop, the loop deviation is determined. When the wall temperature deviation, the flue gas temperature deviation, and the loop deviation all exceed the corresponding preset thresholds in multiple consecutive sampling periods, it is determined that the heat-receiving surface damage is suspected in the corresponding analysis area, and the short-term diagnostic window is established.
[0008] Furthermore, the key operating parameters of the boiler are kept within a preset fluctuation range, and all diagnostic actions are executed starting from the same target baseline range, specifically including: The main load, total air volume, total coal feed, and total desuperheating water volume are controlled within their respective preset fluctuation ranges. Before performing the next diagnostic action, restore the boiler’s key operating parameters to the same target baseline range as at the start of the previous diagnostic action; Based on the current anomaly level of the target analysis area, the boiler safety boundary, and the minimum distinguishable requirement of the corresponding diagnostic action, the minimum action amplitude that satisfies the amplification difference and does not exceed the safety upper limit is selected from the preset action amplitude table.
[0009] Furthermore, determining the net recovery amount of the target analysis area specifically includes: First, a first short-term soot blowing action is performed on the soot blowing device corresponding to the target analysis area. After the response enters the preset stable judgment interval, a second short-term soot blowing action is performed on the reference area with the same effect to obtain the first recovery amount and the first accompanying change amount. After the response is restored to the target baseline range, a third short-term soot blowing action is first performed on the same reference area. After the response enters the preset stable judgment interval, a fourth short-term soot blowing action is performed on the target analysis area to obtain the second recovery amount and the second accompanying change amount. The net recovery amount anchored to the target analysis region is determined according to the following formula: ;in, Indicates net recovery amount. This represents the first recovery amount of the target analysis area after the first short-term soot blowing action. This represents the first accompanying change in the target analysis region caused by the aforementioned reference region during the execution of the second short-term soot blowing action. This represents the second recovery amount of the target analysis area after the fourth short-term soot blowing action. This represents the second accompanying change in the target analysis region caused by the aforementioned co-operating reference region when the third short-term soot blowing action is performed.
[0010] Furthermore, determining the migration amount of the anomaly center specifically includes: After returning to the target baseline range, the coal feed rate of the burner group on one side of the target analysis area is increased while the coal feed rate of the burner group on the other side is decreased simultaneously to perform local combustion distribution adjustment in the first direction; After returning to the target baseline range, the coal feed rate of the burner group on one side of the target analysis area is reduced while the coal feed rate of the burner group on the other side is increased simultaneously to perform a local combustion distribution adjustment in the second direction. The thermal deviation distribution of the target analysis area, the symmetrical reference area, and the adjacent analysis area under two local combustion distribution adjustments was collected. The abnormal center positions corresponding to the two local combustion distribution adjustments are determined respectively, and the migration amount of the abnormal center is determined based on the displacement between the two abnormal center positions.
[0011] Furthermore, determining the relative hysteresis of the target analysis region relative to the same loop reference region specifically includes: After returning to the target baseline range, perform a small load increase action with a first preset amplitude; After returning to the target baseline range, perform a small load reduction action equal to the first preset amplitude; Collect wall temperature data of the target analysis area and the same loop reference area under the same load value within the overlapping load range corresponding to the two actions; Determine the difference in rise and fall of the target analysis area under the same load value; Determine the median value of the rise and fall difference of the reference area in the same circuit under the same load value; The relative hysteresis is determined based on the median value of the rise and fall difference of the target analysis area under the same load value and the rise and fall difference of the reference area in the same loop under the same load value.
[0012] Furthermore, the retention of preset damage types specifically includes: A pre-defined set of damage types is established, including smoke-side attachments, combustion deviation, local flow anomalies, local insufficient cooling, metal degradation, and measuring point anomalies. When the net recovery exceeds the first threshold, the smoke-side deposits are retained. When the migration amount of the abnormal center exceeds the second threshold, the combustion deflection class is retained; When the relative hysteresis exceeds the third threshold, the categories of local flow anomaly, local insufficient cooling, and metal degradation are retained. When a stable response pattern is not formed under short-term soot blowing, local combustion distribution adjustment, and small load increase and decrease actions, and the anomaly is concentrated only at a single measuring point, the aforementioned measuring point anomaly type is retained; The preset damage types to be retained form a candidate set.
[0013] Furthermore, in step S5, deleting preset damage types from the candidate set and determining the target damage type specifically includes: Set corresponding deletion conditions for each preset damage type; When any preset damage type meets the corresponding deletion condition under both sets of diagnostic actions, the preset damage type is deleted from the candidate set. When one preset damage type remains in the candidate set, the remaining preset damage type is determined as the target damage type. When two or more preset damage types remain in the candidate set, the primary damage type and secondary damage type are determined according to the preset priority, and the corresponding analysis results are output.
[0014] This invention also provides an intelligent analysis system for boiler heating surface damage, comprising: The analysis zone construction unit is used to divide the heating surface of the target boiler into multiple analysis zones, and to establish a symmetrical reference zone, a co-function reference zone, and a co-loop reference zone for each analysis zone; The status data acquisition unit is used to acquire the status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. The net recovery determination unit controls the key operating parameters of the boiler to be within a preset fluctuation range within the short-term diagnostic window, and makes each diagnostic action start from the same target baseline range; it performs two short-term soot blowing actions with alternating order on the target analysis area and its corresponding reference area to determine the net recovery of the target analysis area. The migration and hysteresis determination unit, after recovering to the target baseline range, performs two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled to the target analysis area to determine the migration of the abnormal center; after recovering to the target baseline range, it performs small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis of the target analysis area relative to the reference area in the same loop. The damage type determination unit, based on the net recovery amount, the abnormal center migration amount, and the relative backlash amount, retains or deletes preset damage types, determines the target damage type, and outputs the corresponding analysis results.
[0015] (III) Beneficial Effects Compared with the prior art, the present invention provides an intelligent analysis method and system for boiler heating surface damage, which has the following beneficial effects: 1. This intelligent analysis method and system for boiler heating surface damage divides the boiler heating surface into multiple analysis zones and establishes symmetrical reference zones, co-operation reference zones, and co-loop reference zones respectively. When suspected heating surface damage points appear in the target analysis zone, a short-term diagnostic window is established, and all diagnostic actions are initiated within the same target baseline range. This establishes the analysis of different regions and actions on a unified reference relationship and unified operating condition basis, solving the problems of existing technologies that rely on single measurement point over-limit, static threshold, or single trend analysis, which lead to difficulty in regional comparison, inaccurate anomaly location, and insufficient stability of analysis results under complex operating conditions.
[0016] 2. This intelligent analysis method and system for boiler heating surface damage determines the net recovery amount by executing two short-term soot blowing actions with alternating sequences, determines the abnormal center migration amount by executing two local combustion distribution adjustments with opposite directions and equal amplitudes, and determines the relative hysteresis amount by executing small load increase and small load decrease actions with equal amplitudes. Based on the net recovery amount, abnormal center migration amount, and relative hysteresis amount, preset damage types are retained or deleted. It can effectively distinguish between smoke-side attachments, combustion deviation, local flow abnormalities, local insufficient cooling, metal degradation, and measuring point abnormalities. It solves the problems in existing technologies where different abnormal sources are easily confused, damage types are not clearly identified, and the accuracy of analysis results is insufficient under coupled operating conditions such as soot blowing disturbances, load fluctuations, and combustion adjustments. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating the intelligent analysis method for boiler heating surface damage provided by the present invention. Figure 2 This is a schematic diagram of the intelligent analysis system for boiler heating surface damage provided by the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0020] Please see Figure 1 , Figure 1 A flowchart illustrating the intelligent analysis method for boiler heating surface damage provided by this invention; the intelligent analysis method for boiler heating surface damage includes: S1. Divide the heating surface of the target boiler into multiple analysis zones, and establish symmetrical reference zones, identical action reference zones, and identical loop reference zones for each analysis zone; Specifically, the spatial range of the target boiler's heating surface is first regionalized, and then a fixed reference object is configured for each analysis zone. During implementation, the boiler heating surface layout diagram, header connection relationships, parallel branch relationships, and sootblower coverage area are read first. The tube bank area involved in damage analysis within the heating surface is then expanded into a planar analysis interface, and multiple analysis zones are formed according to predetermined rules. Each analysis zone corresponds to a continuous heating surface area, containing continuously positioned tube banks or tube panels. After the analysis zones are delineated, structural symmetry reference relationships, reference relationships for the same sootblowing action, and reference relationships for the same medium loop are established respectively. This ensures that subsequent comparisons of state data are no longer based on single-point measurements, but rather on the differences between the target analysis zone and the corresponding reference zone.
[0021] Furthermore, in one embodiment provided in this application, S1 specifically includes: The target boiler's heating surface is divided into multiple analysis zones according to the medium flow path; For each analysis region, an analysis region that is symmetrical to its position in the boiler structure is identified as a symmetrical reference region; The analysis area subjected to the same soot blowing device is identified as the reference area for the same action; The analysis area located in the same header and the same parallel branch is identified as the reference area of the same loop.
[0022] Specifically, when dividing the analysis zone according to the medium flow path, the heated surface is divided into several layers along the medium inlet to outlet direction. Then, combined with the transverse tube bank distribution, each layer is further divided into multiple local regions, thus forming a two-dimensional analysis grid. Taking water-cooled walls, superheaters, or reheaters as examples, the longitudinal layering is first determined based on the inlet header, outlet header, and intermediate parallel branch routing. Then, the transverse division is determined based on the tube bank distribution along the furnace width direction, ultimately resulting in multiple analysis zones with fixed boundaries. Each analysis zone can be assigned a unique number during implementation; this number is used to bind status data such as wall temperature, flue gas temperature, steam temperature, flow rate, and pressure.
[0023] When determining the symmetrical reference region, the centerline of the boiler heating surface is used as the axis of symmetry. On the other side of the centerline, an analysis region with the same lateral distance and consistent longitudinal height is selected as the symmetrical reference region. When the target analysis region is located near the center, the nearest corresponding region on the other side of the centerline is selected as the symmetrical reference region. This reference relationship is used to characterize local differences under conditions of consistent structural positions, ensuring comparability between the target analysis region and its corresponding region on the other side.
[0024] When determining the reference zone for the same action, based on the sootblower's installation location, the direction of the sootblowing medium injection, and the coverage trajectory, first mark the corresponding action area for each sootblower device. Then, select an analysis area within the same action area that differs from the target analysis area as the reference zone for the same action. For example, when a long telescopic sootblower covers multiple adjacent analysis areas in the same layer, the area within these analysis areas that is on the same sootblowing path as the target analysis area can be selected as the reference zone for the same action. This reference relationship is used to distinguish between changes in the target analysis area itself and accompanying changes caused by sootblowing disturbances before and after the sootblowing action.
[0025] When determining the reference zone for the same loop, based on the piping connections of the boiler steam-water system or steam system, trace the inlet header, outlet header, and intermediate parallel branches corresponding to the heating surface where the target analysis zone is located, and select the reference zone from other analysis zones located in the same header and the same parallel branch. For multiple analysis zones connected in series along the direction of medium flow, the analysis zone adjacent to the target analysis zone in the flow direction is preferentially selected as the reference zone for the same loop. This reference relationship is used to compare the differences in thermal response of different regions within the same loop during load increase and decrease.
[0026] S2. Obtain the status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. Specifically, the process begins with acquiring and aligning status data over time. Then, deviations are calculated based on different reference relationships, and suspected damage to the heated surface is identified using continuous sampling results. During implementation, wall temperature, flue gas temperature, steam temperature, flow rate, and pressure measurement points for each analysis zone are connected to the same data acquisition platform. Data is collected according to a unified sampling period, and the timestamps are aligned. For analysis zones with multiple similar measurement points, the data from these similar points is aggregated to form the status data for that zone. Wall temperature deviation, flue gas temperature deviation, and loop deviation are then calculated. Based on the deviation changes over multiple consecutive sampling periods, it is determined whether the analysis zone has entered an abnormal state. When the abnormality criteria are met for multiple consecutive sampling periods, the analysis zone is designated as the target analysis zone, and a short-term diagnostic window is established starting from the moment the abnormality criteria are first met.
[0027] Furthermore, in one embodiment provided in this application, S2 specifically includes: Obtain wall temperature data, flue gas temperature data, steam temperature data, flow rate data, and pressure data for each analysis zone; Based on the wall temperature data corresponding to each analysis zone and the wall temperature data corresponding to the symmetrical reference zone, the wall temperature deviation is determined. Based on the smoke temperature data corresponding to each analysis zone and the smoke temperature data corresponding to the reference zone with the same effect, the smoke temperature deviation is determined; Based on the steam temperature, flow rate, and pressure data corresponding to each analysis zone and the steam temperature, flow rate, and pressure data corresponding to the reference zone in the same loop, the loop deviation is determined. When the wall temperature deviation, flue gas temperature deviation, and loop deviation all exceed the corresponding preset thresholds in multiple consecutive sampling periods, suspected damage to the heated surface in the corresponding analysis area is identified, and a short-term diagnostic window is established.
[0028] Specifically, when acquiring status data, synchronous data is collected for each analysis zone according to a fixed sampling period. For wall temperature data, metal wall temperature measuring points located on the outer surface of the pipe wall are used; for flue gas temperature data, flue gas temperature measuring points located at corresponding positions in the furnace or flue can be used; for steam temperature, flow rate, and pressure data, process measuring points at the header outlet, branch inlet, or branch outlet can be used. When multiple similar measuring points exist within an analysis zone, the average value of each measuring point within that analysis zone in the current sampling period is first calculated as the regional representative value for that analysis zone in the current sampling period. If an invalid value appears at a measuring point in the current sampling period, the invalid value can be removed before calculating the regional representative value.
[0029] The wall temperature deviation is determined by the difference between the representative wall temperature value of the target analysis area and the representative wall temperature value of the symmetrical reference area. Let the target analysis area be in the [missing information - likely a specific region or area]. The wall temperature represented by each sampling period is The symmetric reference region is in the first The wall temperature represented by each sampling period is Then the first The wall temperature deviation for each sampling period can be expressed as: ;when When the value is positive and continues to increase, it indicates that there is a local temperature rise that is too high in the target analysis area relative to the symmetrical position of the structure.
[0030] The smoke temperature deviation is determined by the difference between the representative smoke temperature value of the target analysis area and the representative smoke temperature value of the reference area under the same action. Let the target analysis area be in the [missing information - likely a specific region or area]. The representative value of smoke temperature for each sampling period is The same reference area in the first The representative value of smoke temperature for each sampling period is Then the first The smoke temperature deviation for each sampling period can be expressed as: This deviation is used to characterize the difference in the thermal state of the smoke side between different analysis zones within the same soot blowing range.
[0031] The loop deviation is determined based on a combination of steam temperature difference, flow rate difference, and pressure difference. During implementation, the target analysis zone and the reference zone in the same loop can be calculated first. Steam temperature difference within each sampling period Poor flow and pressure difference Then, the values are weighted and summed according to preset weights to obtain the loop deviation. Let the weighting coefficients be respectively... Then the first The loop deviation for each sampling period can be expressed as: ;in, It is preset based on boiler type, branch characteristics, and historical operating data. This combined quantity can simultaneously incorporate deviations in thermal state and flow state within the same medium circuit into the judgment process.
[0032] When identifying suspected damage to the heated surface, set wall temperature thresholds for wall temperature deviation, flue gas temperature deviation, and loop deviation, respectively. Smoke temperature threshold and loop threshold And set the number of consecutive judgment cycles. For any analysis region, when starting from the first... Starting from each sampling period, continuously All sampling periods satisfy , and When this occurs, the analysis area is identified as potentially showing signs of heat-receiving surface damage. The target analysis area is directly selected from the analysis area that first meets the continuous judgment criteria; the short-term diagnostic window is established from the start of the sampling cycle when the criteria are first met and covers the preset duration thereafter, for use by subsequent short-term soot blowing operations, local combustion distribution adjustments, and load increases / decreases.
[0033] S3. Within the short-term diagnostic window, control the key operating parameters of the boiler to be within the preset fluctuation range, and ensure that each diagnostic action starts from the same target baseline range; perform two short-term soot blowing actions with alternating order on the target analysis area and its corresponding reference area to determine the net recovery amount of the target analysis area; Specifically, after the short-term diagnostic window is established, the overall boiler operating condition is first constrained, and then diagnostic actions are performed for the target analysis region. During implementation, the main load, total air volume, total coal feed, and total desuperheating water volume are used as key operating parameters. Within the short-term diagnostic window, the deviation between the current value and the baseline value is continuously collected and compared in real time. When any key operating parameter exceeds the corresponding allowable deviation, the next diagnostic action is paused, and execution continues only after the key operating parameter returns to the target baseline range. In this way, the response changes of the target analysis region before and after different diagnostic actions are all based on the same operating condition, and the subsequent recovery amount, accompanying change amount, and net recovery amount are comparable.
[0034] Furthermore, in one embodiment provided in this application, controlling the key operating parameters of the boiler within a preset fluctuation range and ensuring that each diagnostic action is executed starting from the same target baseline range specifically includes: The main load, total air volume, total coal feed, and total desuperheating water volume are controlled within their respective preset fluctuation ranges. Before performing the next diagnostic action, restore the boiler’s key operating parameters to the same target baseline range as at the start of the previous diagnostic action; Based on the current anomaly level of the target analysis area, the boiler safety boundary, and the minimum distinguishable requirement of the corresponding diagnostic action, the minimum action amplitude that meets the amplification difference and does not exceed the safety upper limit is selected from the preset action amplitude table.
[0035] Specifically, the target baseline range is the continuous range before the start of the short-term diagnostic window. The average range of main load, total air volume, total coal feed and total desuperheating water volume within each sampling period; A value of 5, 8, or 10 can be selected. During implementation, allowable deviations for the main load should be set accordingly. Total air volume allowable deviation Total coal feed allowable deviation Allowable deviation of total desuperheating water volume As long as the current value satisfies: , , , This allows us to determine if the current operating condition has returned to the target baseline range. These represent the baseline values for the main load, total air volume, total coal feed, and total desuperheating water volume corresponding to the target baseline range. A preset action amplitude table can be pre-stored in the control system, and the corresponding action amplitudes are recorded according to the degree of anomaly. For example, multiple soot blowing duration settings can be set for short-term soot blowing actions, multiple coal feed adjustment settings can be set for subsequent combustion distribution adjustments, and multiple load change settings can be set for load increase and decrease actions. During execution, the classification interval corresponding to the current degree of anomaly in the target analysis area is first read, and then the minimum action amplitude under that classification is retrieved from the preset action amplitude table for this round of diagnostic actions.
[0036] Furthermore, in one embodiment provided in this application, determining the net recovery of the target analysis region specifically includes: First, perform a first short-term soot blowing action on the soot blowing device corresponding to the target analysis area. After the response enters the preset stable judgment interval, perform a second short-term soot blowing action on the reference area with the same effect to obtain the first recovery amount and the first accompanying change amount. After the response is restored to the target baseline range, a third short-term soot blowing action is performed on the reference area with the same effect. After the response enters the preset stability judgment interval, a fourth short-term soot blowing action is performed on the target analysis area to obtain the second recovery amount and the second accompanying change amount. The net recovery anchored to the target analysis region is determined using the following formula: ;in, Indicates net recovery amount. This represents the initial recovery amount of the target analysis region after the first short-duration dust removal action. This represents the first accompanying change in the target analysis area caused by the reference area during the second short-duration soot blowing action. This represents the second recovery amount of the target analysis region after the fourth short-duration soot blowing action. This represents the second accompanying change in the target analysis area caused by the same action reference area when the third short-term soot blowing action is performed.
[0037] Specifically, the net recovery amount is determined based on the change in the wall temperature deviation of the target analysis zone. Taking the first short-duration soot blowing action as an example, the current wall temperature deviation of the target analysis zone is recorded before the action begins. After the response enters the stable judgment interval, record the wall temperature deviation of the target analysis area. Then the first recovery amount is taken Then, a second short-duration soot blowing action is performed, and the wall temperature deviation of the target analysis area is recorded before the start of the second short-duration soot blowing action. After the response enters the stable judgment interval, record the wall temperature deviation of the target analysis area. Then the first accompanying change is taken The second accompanying change can be obtained from the third and fourth short-duration soot blowing actions in the same way. Second recovery amount .
[0038] Stability judgment interval based on continuity The absolute value of the rate of change of the wall temperature deviation in the target analysis area within each sampling period is less than the preset slope threshold. To confirm; 3 or 5 can be chosen. It can be preset according to the sampling period and measurement point resolution. By reversing the order of blowing the target analysis area first and then the reference area with the same effect, and then blowing the reference area with the same effect first and then the target analysis area, the common disturbances caused by the sootblower action can be separated from the recovery changes of the target analysis area itself. The final net recovery amount is obtained. When the value is positive, it indicates that the target analysis area shows stable recovery after two rounds of alternating soot blowing. The larger the value, the more significant the recoverable changes in the target analysis area before and after the soot blowing action.
[0039] S4. After returning to the target baseline range, perform two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled with the target analysis area to determine the migration amount of the abnormal center; after returning to the target baseline range, perform small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis amount of the target analysis area relative to the reference area in the same loop. Specifically, after completing the short-term soot blowing diagnosis, two types of controlled disturbances are introduced: local combustion distribution disturbance and load increase / decrease disturbance. The local combustion distribution disturbance is used to observe whether the abnormal heat deviation distribution shifts in position as the direction of heat input on the combustion side changes; the load increase / decrease disturbance is used to observe the difference in thermal response between the target analysis area and the reference area in the same loop under the same load value. Before execution, the main load, total air volume, total coal feed, and total desuperheating water volume are restored to the target baseline range. Then, two local combustion distribution adjustments in opposite directions, one load increase action, and one load decrease action are executed according to preset amplitudes. After the two types of disturbances are executed, the abnormal center migration amount and relative hysteresis amount are obtained, respectively, for subsequent damage type screening.
[0040] Furthermore, in one embodiment provided in this application, determining the migration amount of the abnormal center specifically includes: After returning to the target baseline range, increase the coal feed rate of the burner group on one side of the target analysis area and simultaneously decrease the coal feed rate of the burner group on the other side to perform local combustion distribution adjustment in the first direction; After returning to the target baseline range, reduce the coal feed rate of the burner group on one side of the target analysis area and simultaneously increase the coal feed rate of the burner group on the other side to perform local combustion distribution adjustment in the second direction; The thermal deviation distribution of the target analysis area, the symmetrical reference area, and the adjacent analysis areas were collected after two local combustion distribution adjustments. The abnormal center positions corresponding to the two local combustion distribution adjustments are determined respectively, and the migration amount of the abnormal center is determined based on the displacement between the two abnormal center positions.
[0041] Specifically, the local combustion distribution adjustments in the first and second directions use equal coal feed rate adjustment amplitudes. Let the increase in coal feed rate for the burner group on one side of the target analysis zone in the first direction be [value missing]. The coal feed rate of the other burner group decreased by an amount of Let the reduction in coal feed rate of the burner group on one side of the target analysis zone in the second direction be denoted as . The increase in coal feed rate for the other burner group was [value missing]. ,in, After adjustment, the thermal deviation distribution of the target analysis area, the symmetrical reference area, and adjacent analysis areas is continuously collected within the preset observation period, and the thermal deviation of each analysis area is mapped to the center coordinates of the corresponding analysis area.
[0042] The location of the anomaly center is determined using a weighted center method. Let the anomaly center be the one participating in the calculation. The center coordinates of each analysis region are The absolute value of the corresponding thermal deviation is The abnormal center location under a certain local combustion distribution adjustment satisfy: ; ;in, This indicates that the summation is performed on each analysis region involved in the calculation. The anomaly center location corresponding to the first direction is denoted as... The anomaly center location corresponding to the second direction is denoted as Then the migration amount of the abnormal center satisfy: ,in, Indicates the migration amount of the abnormal center. When When the value increases, it indicates that the abnormal heat deviation distribution has shifted position as the combustion distribution direction changes; when... When the value decreases, it indicates that the abnormal thermal deviation distribution remains in a similar region under two opposite adjustments.
[0043] Furthermore, in one embodiment provided in this application, determining the relative hysteresis of the target analysis region relative to the reference region in the same loop specifically includes: After returning to the target baseline range, perform a small load increase action with the first preset amplitude; After returning to the target baseline range, perform a small load reduction action equal to the first preset amplitude; Collect wall temperature data of the target analysis area and the reference area in the same loop under the same load value within the overlapping load range corresponding to the two actions; Determine the difference in load increase or decrease for the target analysis area under the same load value; Determine the median value of the rise / fall difference for the reference zone in the same circuit under the same load value; The relative hysteresis is determined by comparing the difference between the target analysis area and the reference area in the same circuit under the same load value.
[0044] Specifically, let the baseline load be... The first preset amplitude is The load increase action will then transfer the load from Adjust to The load reduction action will change the load from Adjust The overlapping load intervals corresponding to the load increase and load decrease processes are as follows: Select the same load value within this range. Record the wall temperature values of the target analysis area under the load increase process. Wall temperature during load reduction process Then the target analysis area is at the load value The difference between the rise and fall at the point satisfies: .
[0045] For the first in the same circuit Each reference zone, at the same load value Record the wall temperature values at each load increase point. and the wall temperature value under load Then the reference area is at the load value The difference between the rise and fall at the point satisfies: .
[0046] All reference areas on the same circuit are at the load value The median value of the difference between the rise and fall at point is obtained. The target analysis area is at the load value. Relative hysteresis at the location satisfy: .
[0047] If the overlapping load interval is selected For the same load value, the overall relative hysteresis can be determined by the following formula: . It reflects the degree of separation of the wall temperature path of the target analysis area relative to the reference area in the same loop during the load increase and decrease process, and is subsequently used directly as the input for damage type retention and deletion.
[0048] S5. Based on the net recovery amount, the migration amount of the abnormal center, and the relative backlash, retain or delete the preset damage types, determine the target damage type, and output the corresponding analysis results.
[0049] Specifically, the system first screens various damage types based on net recovery, abnormal center migration, and relative hysteresis. Then, it further eliminates candidate types from the initial screening to obtain the target damage type. During implementation, a correspondence between damage types and discriminant parameters is established. Net recovery corresponds to changes in smoke-side adhesion, abnormal center migration corresponds to thermal deviation migration on the combustion side, and relative hysteresis corresponds to thermal response lag within the loop. After calculating the three discriminant parameters, damage types that meet the criteria are first screened according to their respective thresholds. Then, types that do not meet the criteria are removed from the candidate set according to deletion conditions. If only one type remains in the candidate set, it is directly used as the result of this round of analysis. If multiple types remain in the candidate set, the primary and secondary types are output according to preset priorities.
[0050] Furthermore, in one embodiment provided in this application, the preset damage type is retained, specifically including: A pre-defined set of damage types is established, including smoke-side attachments, combustion deviation, local flow anomalies, local insufficient cooling, metal degradation, and measuring point anomalies. When the net recovery exceeds the first threshold, retain the smoke-side attachments. When the migration amount of the abnormal center exceeds the second threshold, the combustion deflection class is retained; When the relative hysteresis exceeds the third threshold, the categories of local flow anomaly, local insufficient cooling, and metal deterioration are retained. When a stable response pattern is not formed under short-term soot blowing, local combustion distribution adjustment, and small load increase and decrease actions, and the anomaly is concentrated only at a single measuring point, the measuring point anomaly category is retained; The preset damage types to be retained form a candidate set.
[0051] Specifically, the candidate set is initialized to an empty set, and screening is performed in a fixed order. Let the net recovery threshold be... The threshold for abnormal center migration is The relative hysteresis threshold is If the net recovery amount Then add the smoke-side attachments to the candidate set; if the abnormal center migration amount Then add the combustion skew class to the candidate set; if the relative hysteresis If local flow anomalies, insufficient local cooling, and metal degradation are identified, then these categories will be added to the candidate set. The screening of measurement point anomalies can employ a single-point independent judgment method: when only one wall temperature measurement point in the target analysis area continuously deviates, while other wall temperature measurement points in the same analysis area, corresponding measurement points in the symmetrical reference area, and corresponding measurement points in the same loop reference area do not show synchronous deviations, and the net recovery, anomaly center migration, and relative hysteresis have not reached their respective thresholds, then the measurement point anomaly category will be added to the candidate set. After all screenings are completed, each type in the candidate set will correspond to at least one discriminant that reaches a threshold.
[0052] Furthermore, in one embodiment provided by this application, step S5, which involves deleting preset damage types from the candidate set and determining the target damage type, specifically includes: Set corresponding deletion conditions for each preset damage type; When any preset damage type meets the corresponding deletion condition under both sets of diagnostic actions, the preset damage type is deleted from the candidate set. When one preset damage type remains in the candidate set, the remaining preset damage type is determined as the target damage type. When there are two or more preset damage types remaining in the candidate set, the primary damage type and secondary damage type are determined according to the preset priority, and the corresponding analysis results are output.
[0053] Specifically, separate deletion condition tables are set up according to damage type. For smoke-side attachments, the deletion condition is set to a net recovery amount not exceeding [amount missing]. And the migration amount of the abnormal center is greater than The deletion condition for combustion deviation classes is set to an anomaly center migration amount not exceeding [a certain value]. And the relative hysteresis is greater than The deletion criteria for categories such as localized flow anomalies, insufficient localized cooling, and metal degradation can be set to a relative hysteresis of no more than [amount missing]. And the net recovery amount is greater than The deletion criteria for abnormal measurement points can be set as two or more adjacent measurement points within the target analysis area deviating synchronously, or deviations in the same direction occurring in symmetrical reference areas or co-loop reference areas. During implementation, first check the response results of each type within the candidate set during short-term soot blowing, localized combustion distribution adjustment, slight load increase, and slight load decrease actions, then count the number of action groups that meet the deletion criteria; if the number of action groups reaches two, delete it directly from the candidate set.
[0054] If only one type remains in the candidate set after deletion, that type is the target damage type. If two or more types remain in the candidate set after deletion, their priority can be determined in the following order: "metal degradation, insufficient local cooling, abnormal local flow, smoke-side deposits, combustion deviation, and abnormal measuring point." The type with the highest priority is designated as the primary type, and the type with the highest priority among the remaining types is designated as the secondary type. The analysis results should include at least the target analysis area number, target damage type, primary type, secondary type, and corresponding discriminant value.
[0055] Please see Figure 2 , Figure 2 This is a schematic diagram of the intelligent analysis system for boiler heating surface damage provided by the present invention; the present invention also provides an intelligent analysis system for boiler heating surface damage, comprising: The analysis zone construction unit is used to divide the heating surface of the target boiler into multiple analysis zones, and to establish symmetrical reference zones, identical action reference zones, and identical loop reference zones for each analysis zone; The status data acquisition unit is used to acquire status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. The net recovery determination unit controls the key operating parameters of the boiler within a preset fluctuation range within a short-term diagnostic window, and ensures that each diagnostic action starts from the same target baseline range; it performs two short-term soot blowing actions with alternating sequences on the target analysis area and its corresponding reference area to determine the net recovery of the target analysis area. The migration and hysteresis determination unit performs two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled with the target analysis area after restoring to the target baseline range, to determine the migration of the abnormal center; after restoring to the target baseline range, it performs small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis of the target analysis area relative to the reference area in the same loop. The damage type determination unit, based on net recovery, anomaly center migration, and relative backlash, retains or deletes preset damage types, determines the target damage type, and outputs the corresponding analysis results.
[0056] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0057] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. An intelligent analysis method for boiler heating surface damage, characterized in that, include: S1. Divide the heating surface of the target boiler into multiple analysis zones, and establish a symmetrical reference zone, a reference zone with the same action, and a reference zone with the same loop for each analysis zone; S1 specifically includes: The target boiler's heating surface is divided into multiple analysis zones according to the medium flow path; For each of the analysis regions, an analysis region that is symmetrical to its position in the boiler structure is determined as the symmetry reference region; The analysis area subjected to the same soot blowing device is identified as the reference area for the same action; The analysis area located in the same header and the same parallel branch is determined as the reference area of the same loop; S2. Obtain the status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. S3. Within the short-term diagnostic window, control the key operating parameters of the boiler to be within the preset fluctuation range, and make each diagnostic action start from the same target baseline range; perform two short-term soot blowing actions with alternating order on the target analysis area and its corresponding reference area to determine the net recovery amount of the target analysis area; S4. After returning to the target baseline range, perform two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled to the target analysis area to determine the abnormal center migration amount; after returning to the target baseline range, perform small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis amount of the target analysis area relative to the same loop reference area. S5. Based on the net recovery amount, the abnormal center migration amount, and the relative backlash amount, retain or delete preset damage types, determine the target damage type, and output the corresponding analysis results.
2. The intelligent analysis method for boiler heating surface damage according to claim 1, characterized in that, S2 specifically includes: Obtain wall temperature data, flue gas temperature data, steam temperature data, flow rate data, and pressure data corresponding to each of the analysis zones; Based on the wall temperature data corresponding to each of the analysis zones and the wall temperature data corresponding to the symmetrical reference zone, the wall temperature deviation is determined. Based on the smoke temperature data corresponding to each analysis zone and the smoke temperature data corresponding to the same action reference zone, the smoke temperature deviation is determined; Based on the steam temperature, flow rate, and pressure data corresponding to each analysis zone and the steam temperature, flow rate, and pressure data corresponding to the reference zone in the same loop, the loop deviation is determined. When the wall temperature deviation, the flue gas temperature deviation, and the loop deviation all exceed the corresponding preset thresholds in multiple consecutive sampling periods, it is determined that the heat-receiving surface damage is suspected in the corresponding analysis area, and the short-term diagnostic window is established.
3. The method for intelligent analysis of boiler heating surface damage according to claim 1, characterized in that, The control of key boiler operating parameters is kept within a preset fluctuation range, and all diagnostic actions are executed starting from the same target baseline range. Specifically, this includes: The main load, total air volume, total coal feed, and total desuperheating water volume are controlled within their respective preset fluctuation ranges. Before performing the next diagnostic action, restore the boiler’s key operating parameters to the same target baseline range as at the start of the previous diagnostic action; Based on the current anomaly level of the target analysis area, the boiler safety boundary, and the minimum distinguishable requirement of the corresponding diagnostic action, the minimum action amplitude that satisfies the amplification difference and does not exceed the safety upper limit is selected from the preset action amplitude table.
4. The method for intelligent analysis of boiler heating surface damage according to claim 1, characterized in that, Determining the net recovery of the target analysis region specifically includes: First, a first short-term soot blowing action is performed on the soot blowing device corresponding to the target analysis area. After the response enters the preset stable judgment interval, a second short-term soot blowing action is performed on the reference area with the same effect to obtain the first recovery amount and the first accompanying change amount. After the response is restored to the target baseline range, a third short-term soot blowing action is first performed on the same reference area. After the response enters the preset stable judgment interval, a fourth short-term soot blowing action is performed on the target analysis area to obtain the second recovery amount and the second accompanying change amount. The net recovery amount anchored to the target analysis region is determined according to the following formula: ;in, Indicates net recovery amount. This represents the first recovery amount of the target analysis area after the first short-term soot blowing action. This represents the first accompanying change in the target analysis region caused by the aforementioned reference region during the execution of the second short-term soot blowing action. This represents the second recovery amount of the target analysis area after the fourth short-term soot blowing action. This represents the second accompanying change in the target analysis region caused by the aforementioned co-operating reference region when the third short-term soot blowing action is performed.
5. The method for intelligent analysis of boiler heating surface damage according to claim 1, characterized in that, The determination of the migration amount of the abnormal center specifically includes: After returning to the target baseline range, the coal feed rate of the burner group on one side of the target analysis area is increased while the coal feed rate of the burner group on the other side is decreased simultaneously to perform local combustion distribution adjustment in the first direction; After returning to the target baseline range, the coal feed rate of the burner group on one side of the target analysis area is reduced while the coal feed rate of the burner group on the other side is increased simultaneously to perform a local combustion distribution adjustment in the second direction. The thermal deviation distribution of the target analysis area, the symmetrical reference area, and the adjacent analysis area under two local combustion distribution adjustments was collected. The abnormal center positions corresponding to the two local combustion distribution adjustments are determined respectively, and the migration amount of the abnormal center is determined based on the displacement between the two abnormal center positions.
6. The method for intelligent analysis of boiler heating surface damage according to claim 1, characterized in that, Determining the relative hysteresis of the target analysis region relative to the reference region in the same loop specifically includes: After returning to the target baseline range, perform a small load increase action with a first preset amplitude; After returning to the target baseline range, perform a small load reduction action equal to the first preset amplitude; Collect wall temperature data of the target analysis area and the same loop reference area under the same load value within the overlapping load range corresponding to the two actions; Determine the difference in rise and fall of the target analysis area under the same load value; Determine the median value of the rise and fall difference of the reference area in the same circuit under the same load value; The relative hysteresis is determined based on the median value of the rise and fall difference of the target analysis area under the same load value and the rise and fall difference of the reference area in the same loop under the same load value.
7. The intelligent analysis method for boiler heating surface damage according to claim 1, characterized in that, The retention of preset damage types specifically includes: A pre-defined set of damage types is established, including smoke-side attachments, combustion deviation, local flow anomalies, local insufficient cooling, metal degradation, and measuring point anomalies. When the net recovery exceeds the first threshold, the smoke-side deposits are retained. When the migration amount of the abnormal center exceeds the second threshold, the combustion deflection class is retained; When the relative hysteresis exceeds the third threshold, the categories of local flow anomaly, local insufficient cooling, and metal degradation are retained. When a stable response pattern is not formed under short-term soot blowing, local combustion distribution adjustment, and small load increase and decrease actions, and the anomaly is concentrated only at a single measuring point, the aforementioned measuring point anomaly type is retained; The preset damage types to be retained form a candidate set.
8. The method for intelligent analysis of boiler heating surface damage according to claim 7, characterized in that, In step S5, deleting preset damage types from the candidate set and determining the target damage type specifically includes: Set corresponding deletion conditions for each preset damage type; When any preset damage type meets the corresponding deletion condition under both sets of diagnostic actions, the preset damage type is deleted from the candidate set. When one preset damage type remains in the candidate set, the remaining preset damage type is determined as the target damage type. When two or more preset damage types remain in the candidate set, the primary damage type and secondary damage type are determined according to the preset priority, and the corresponding analysis results are output.
9. An intelligent analysis system for damage of a boiler heating surface, characterized by, include: The analysis zone construction unit is used to divide the heating surface of the target boiler into multiple analysis zones, and to establish a symmetrical reference zone, a co-function reference zone, and a co-loop reference zone for each analysis zone; The analysis region construction unit specifically includes: The target boiler's heating surface is divided into multiple analysis zones according to the medium flow path; For each of the analysis regions, an analysis region that is symmetrical to its position in the boiler structure is determined as the symmetry reference region; The analysis area subjected to the same soot blowing device is identified as the reference area for the same action; The analysis area located in the same header and the same parallel branch is determined as the reference area of the same loop; The status data acquisition unit is used to acquire the status data of each analysis area. When suspected damage to the heated surface occurs in any analysis area for multiple consecutive sampling cycles, the analysis area with suspected damage to the heated surface is identified as the target analysis area, and a short-term diagnostic window is established. The net recovery determination unit controls the key operating parameters of the boiler to be within a preset fluctuation range within the short-term diagnostic window, and makes each diagnostic action start from the same target baseline range; it performs two short-term soot blowing actions with alternating order on the target analysis area and its corresponding reference area to determine the net recovery of the target analysis area. The migration and hysteresis determination unit, after recovering to the target baseline range, performs two local combustion distribution adjustments with opposite directions and equal amplitudes on the burner group coupled to the target analysis area to determine the migration of the abnormal center; after recovering to the target baseline range, it performs small load increase and small load decrease actions with equal amplitudes to determine the relative hysteresis of the target analysis area relative to the reference area in the same loop. The damage type determination unit, based on the net recovery amount, the abnormal center migration amount, and the relative backlash amount, retains or deletes preset damage types, determines the target damage type, and outputs the corresponding analysis results.