A remote intelligent diagnosis method and system for a pulverized coal boiler based on a one-way secure acquisition link

Abstract

The application discloses a kind of coal powder boiler remote intelligent diagnosis method and system based on one-way secure acquisition link, it is related to industrial boiler intelligent operation and maintenance technical field.The method comprises: obtaining boiler operation data from DCS, after OPC communication, UDP encapsulation, one-way isolation net lock positive transmission and MQTT topic forwarding, upload to cloud platform;Operation data is cleaned, aligned, load interval identification and forms to be diagnosed data set;Respectively execute NO x Diagnosis, air preheater ash deposition grading diagnosis, heating surface coking trend prediction and coal consumption energy efficiency analysis;The fusion of each diagnostic result generates early warning level, disposal priority, optimization suggestion and periodic diagnosis report.The application realizes remote active identification of coal powder boiler emission, ash deposition, coking and energy efficiency problem while ensuring the isolation of industrial control network and external network, reduces the dependence on artificial inspection, improves the timeliness of abnormal discovery and the economy of boiler operation.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology for industrial boilers, specifically to a remote intelligent diagnostic method and system for pulverized coal boilers based on a one-way secure data acquisition link. Background Technology

[0002] Pulverized coal boilers are key steam production equipment in industrial settings such as coal chemical, thermal power, and metallurgy. Their operating status directly affects production continuity, fuel consumption, environmental emissions, and equipment lifespan. Boiler operation involves factors such as load, main steam pressure, main steam temperature, air volume, air temperature, oxygen content, and NO. x Multiple parameters, including emission concentration, ammonia slip concentration, air preheater differential pressure, furnace outlet flue gas temperature, and coal consumption, are strongly coupled, making it difficult for a single measuring point or fixed threshold to accurately reflect the true operational risk of the boiler. Existing boiler monitoring systems mostly rely on on-site DCS or manual inspections. Operators typically adjust ammonia injection valve opening, air volume, and oxygen quantity based on experience, which can easily lead to NO2 emissions. x The system exhibits a "seesaw" effect, with emissions exceeding limits or ammonia escape exceeding limits. For air preheater ash accumulation and heating surface coking, on-site assessments often rely on periodic inspections or differential pressure experience, resulting in delayed early warnings and difficulty in pinpointing specific areas. Coal consumption and energy efficiency analysis typically only provides statistical results, failing to link with diagnostics of abnormalities such as emissions, ash accumulation, and coking. More critically, the boiler DCS is a core system of the industrial control network; direct connection to external networks poses serious security risks. Existing remote monitoring solutions often focus on predicting single problems, lacking a complete solution that unifies emissions, ash accumulation, coking, and energy efficiency analysis into a single diagnostic engine while ensuring unidirectional isolation between the industrial control network and external networks. Therefore, a method and system are urgently needed that can meet the network security isolation requirements of the industrial control network and perform multi-dimensional remote intelligent diagnostics of pulverized coal boilers based on DCS operating data. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the purpose of this invention is to provide a remote intelligent diagnostic method and system for pulverized coal boilers based on a one-way secure acquisition link. This system enables remote and proactive identification of emissions, ash accumulation, coking, and energy efficiency issues in pulverized coal boilers, reducing reliance on manual inspections and improving the timeliness of anomaly detection and the economic efficiency of boiler operation.

[0004] To achieve the above objectives, the embodiments of this invention provide the following technical solutions:

[0005] This application provides a remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link, comprising the following steps: S1, acquiring boiler operation data from the DCS of the pulverized coal boiler via OPC communication service; S2, encapsulating the boiler operation data into UDP packets by the field acquisition service, transmitting the UDP packets in a forward unidirectional manner through a one-way isolation gateway, and converting the UDP packets into MQTT data topics and uploading them to the cloud platform by the external network forwarding service; S3, performing data cleaning, time alignment, outlier processing, measurement point mapping, and load range identification on the boiler operation data on the cloud platform to form a dataset to be diagnosed; S4, executing NO based on the dataset to be diagnosed. x The system integrates ammonia escape optimization diagnosis, air preheater ash accumulation classification diagnosis, heating surface coking trend prediction, and coal consumption efficiency analysis; S5, it merges the diagnostic results to generate early warning levels, handling priorities, operation optimization suggestions, and periodic diagnostic reports.

[0006] Furthermore, the load interval identification in S3 includes: S31, dividing the continuous operating data into multiple load intervals according to the preset load range; S32, establishing emission benchmarks, air preheater differential pressure benchmarks, furnace outlet flue gas temperature trend benchmarks, and coal consumption benchmarks for each load interval; S33, comparing the current operating condition with the benchmark operating condition in the same or similar load intervals during diagnosis.

[0007] Furthermore, NO in S4 x Co-optimization diagnostics with ammonia slip include: S41, obtaining NO within the same load range x Emission concentration, ammonia slip concentration, ammonia injection valve opening, air volume, air temperature, and oxygen content; S42, based on NO x Emission status is determined by emission limits and ammonia slip limits; S43, Filter historical operating data to simultaneously meet the NO... x Candidate operating conditions for emission limits and ammonia slip limits; S44, Based on the candidate operating conditions, generate recommended parameters including air volume, air temperature, oxygen content, and ammonia injection valve opening; S45, When NO x When emissions or ammonia escape exceed limits, corresponding coordinated adjustment suggestions will be provided.

[0008] Furthermore, the coordinated adjustment suggestion in S45 includes: S451, when NO x When the emission concentration exceeds the limit but the ammonia slip concentration does not exceed the limit, priority should be given to generating enhanced adjustment suggestions for ammonia injection valve opening, air volume, or oxygen quantity; S452, when the ammonia slip concentration exceeds the limit and NO x When the emission concentration does not exceed the limit, suggestions for adjusting the ammonia injection valve opening or the air volume ratio should be generated first; S453, when NO x When both emission concentration and ammonia slip concentration exceed the limit, joint optimization suggestions for combustion-side parameters and denitrification-side parameters are generated.

[0009] Further, the air preheater ash accumulation classification diagnosis in S4 includes: S461, obtaining the negative pressure difference on side A and side B of each section of the air preheater, wherein the section includes at least one of the following: high temperature section on the primary air side, medium temperature section on the primary air side, low temperature section on the primary air side, high temperature section on the secondary air side, medium temperature section on the secondary air side, and low temperature section on the secondary air side; S462, within the same load range, calculating the deviation value of the negative pressure difference on side A relative to the reference negative pressure difference on side A and the deviation value of the negative pressure difference on side B relative to the reference negative pressure difference on side B, wherein the reference negative pressure difference on side A and the reference negative pressure difference on side B are derived from the air preheater pressure difference reference; S463, within the same load range, calculating the... S464. Based on the deviation value of side A and the deviation value of side B, the ash accumulation status is divided into normal, light ash accumulation, moderate ash accumulation, and severe ash accumulation; S465. Based on the ash accumulation status, the soot blowing section, soot blowing priority, and suggested handling time are output; S466. When the same section reaches moderate or severe ash accumulation in multiple consecutive load intervals, the soot blowing priority of the corresponding section is increased; S467. When multiple sections in the low temperature section and the medium temperature section of the primary air side, and the low temperature section and the medium temperature section of the secondary air side simultaneously reach moderate or severe ash accumulation, a combined soot blowing suggestion is generated.

[0010] Further, the prediction of coking trend of the heating surface in S4 includes: S471, extracting the time series of flue gas temperature at the furnace outlet within the same load range; S472, performing trend fitting on the time series to obtain the slope of the flue gas temperature at the furnace outlet as a function of operating time; S473, generating a coking warning for the heating surface when the slope exceeds a preset slope threshold or the cumulative increase in flue gas temperature at the furnace outlet exceeds a preset temperature rise threshold; S474, based on the slope or cumulative temperature rise of the flue gas temperature at the furnace outlet, and in conjunction with the boiler operating load, coal consumption, and NO... x The trend of emission concentration or ammonia escape concentration is used to determine the coking risk level; in particular, when the flue gas temperature at the furnace outlet continues to rise and coal consumption rises abnormally or NO... x When the emission concentration or ammonia escape concentration exceeds the limit, the coking risk level is increased.

[0011] Furthermore, the coal consumption energy efficiency analysis in S4 includes: S481, calculating the unit steam coal consumption based on coal consumption and steam output; S482, calculating the operational stability index based on boiler operating load, main steam pressure, and main steam temperature; S483, comparing the unit steam coal consumption and operational stability index with historical or design benchmarks; S484, when the unit steam coal consumption abnormally increases, triggering NO2. xThe results of co-optimization diagnosis of ammonia escape, graded diagnosis of air preheater ash accumulation, and prediction of coking trend of heated surfaces are used to generate an energy efficiency cause analysis.

[0012] Accordingly, this application also provides a remote intelligent diagnostic system for pulverized coal boilers based on a one-way secure acquisition link, comprising: a field data acquisition layer for acquiring boiler operation data from the DCS of the pulverized coal boiler via OPC communication service; a one-way secure transmission layer connected to the field data acquisition layer for encapsulating the boiler operation data into UDP packets by the field acquisition service, enabling the UDP packets to be transmitted unidirectionally in the forward direction via a one-way isolation gateway, and having the UDP packets converted into MQTT data topics and uploaded to the cloud platform by an external network-side forwarding service; a cloud preprocessing layer connected to the one-way secure transmission layer for performing data cleaning, time alignment, outlier processing, measurement point mapping, and load range identification on the boiler operation data on the cloud platform to form a dataset to be diagnosed; and a multi-module diagnostic layer connected to the cloud preprocessing layer for performing NO on the dataset to be diagnosed. x The system integrates ammonia escape optimization diagnosis, air preheater ash accumulation classification diagnosis, heating surface coking trend prediction, and coal consumption energy efficiency analysis; the diagnostic fusion output layer, connected to the multi-module diagnostic layer, is used to fuse the diagnostic results to generate early warning levels, handling priorities, operation optimization suggestions, and periodic diagnostic reports.

[0013] Furthermore, the multi-module diagnostic layer includes: an emission synergistic optimization module, used to perform the NO... x The system includes a collaborative optimization diagnostic module for ammonia slip; an ash accumulation diagnostic module for performing ash accumulation classification diagnostics on the air preheater; a coking prediction module for predicting the coking trend of the heating surface; and an energy efficiency analysis module for performing coal consumption energy efficiency analysis.

[0014] Furthermore, the one-way isolation gateway in the one-way secure transmission layer is configured to only allow forward data transmission from the industrial control network side to the external network side, and the external network-side forwarding service in the one-way secure transmission layer is configured to generate MQTT topics or message fields according to the device number, boiler number, measurement point type, and timestamp.

[0015] The beneficial effects of this invention are as follows: By physically and logically blocking the reverse access path of the external network, the risk of industrial control network security is completely eliminated. At the same time, through standardized data preprocessing and load range division, the diagnostic deviation caused by load fluctuations is avoided from the data source, improving the accuracy of diagnostic results. The parallel implementation of four major categories of special diagnostics can cover all dimensions of environmental protection, equipment, and energy efficiency operation issues. The unified integration and processing of multi-module diagnostic results can also output graded early warning, handling priority, executable optimization suggestions and periodic reports, which greatly reduces the reliance on manual operation and maintenance and the operational threshold. Attached Figure Description

[0016] Figure 1 A flowchart illustrating a remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure data acquisition link, provided in an embodiment of this application;

[0017] Figure 2 This is a schematic diagram of the structure of a remote intelligent diagnostic system for pulverized coal boilers based on a one-way secure acquisition link, provided in an embodiment of this application. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] In this invention, the terms "system" and "network" are used interchangeably. "Multiple" refers to two or more; therefore, in this invention, "multiple" can also be understood as "at least two." "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, it should be understood that in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0020] Example 1:

[0021] like Figure 1 As shown, this application provides a remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link, including the following steps: S1, acquiring boiler operation data from the DCS of the pulverized coal boiler via OPC communication service; S2, encapsulating the boiler operation data into UDP packets by the field acquisition service, transmitting the UDP packets in a forward unidirectional manner through a one-way isolation gateway, and converting the UDP packets into MQTT data topics and uploading them to the cloud platform by the external network forwarding service; S3, performing data cleaning, time alignment, outlier processing, measurement point mapping, and load range identification on the boiler operation data on the cloud platform to form a dataset to be diagnosed; S4, executing NO based on the dataset to be diagnosed. x The system integrates ammonia escape optimization diagnosis, air preheater ash accumulation classification diagnosis, heating surface coking trend prediction, and coal consumption efficiency analysis; S5, it merges the diagnostic results to generate early warning levels, handling priorities, operation optimization suggestions, and periodic diagnostic reports.

[0022] In some embodiments, the OPC communication service deployed on the field industrial control computer is first used to read data from the pulverized coal boiler DCS system, including unit load, main steam pressure, main steam temperature, feedwater flow rate, coal consumption, air volume, air temperature, oxygen content, and NO. x The system collects all operational data, including emission concentration, ammonia slip concentration, ammonia injection valve opening, negative pressure difference at various measuring points of the air preheater, and flue gas temperature at the furnace outlet. The field data acquisition service then encapsulates the raw data from multiple measuring points into lightweight UDP packets with a unified timestamp. These UDP packets are then sent to a one-way isolation gateway, enabling forward unidirectional transmission from the industrial control network to the external network via a hardware-level physical channel. Chip-level blocking of all reverse data packets is achieved. The external network forwarding service receives and parses the UDP packets, converting them into IoT-compatible MQTT data topics and uploading them to the cloud platform. Upon receiving the data, the cloud platform first writes it to a dedicated time-series database for persistent storage. Then, it sequentially performs data cleaning to remove disconnected, over-range, and null data; time alignment to unify the sampling timestamps of all measuring points; outlier processing to filter transient data jumps; and measuring point mapping to convert DCS custom measuring point codes into platform standard codes and identify load intervals. Finally, a standardized diagnostic dataset is generated, and NO is then synchronously initiated based on this dataset. x The system integrates four independent computational processes: ammonia slip co-optimization diagnosis, air preheater ash accumulation classification diagnosis, heating surface coking trend prediction, and coal consumption energy efficiency analysis. Finally, it merges the original diagnostic results output by all modules to generate a unified warning level, handling priority, textual operation optimization suggestions, and periodic diagnostic reports.

[0023] By physically and logically blocking reverse access paths to the external network, the risks to industrial control network security are completely eliminated. At the same time, through standardized data preprocessing and load range division, diagnostic biases caused by load fluctuations are avoided from the data source, improving the accuracy of diagnostic results. The parallel implementation of four major categories of specialized diagnostics can cover all dimensions of environmental protection, equipment, and energy efficiency. The unified integration and processing of multi-module diagnostic results can also output graded early warnings, handling priorities, executable optimization suggestions, and periodic reports, significantly reducing the reliance on manual operation and maintenance and the operational threshold.

[0024] In this embodiment of the application, the load interval identification in S3 includes: S31, dividing the continuous operating data into multiple load intervals according to the preset load range; S32, establishing emission benchmarks, air preheater differential pressure benchmarks, furnace outlet flue gas temperature trend benchmarks, and coal consumption benchmarks for each load interval; S33, comparing the current operating condition with the benchmark operating condition in the same or similar load intervals during diagnosis.

[0025] In some embodiments, multiple continuous load range thresholds are first pre-configured on the cloud platform. The continuous boiler operation data received from the cloud is automatically divided into corresponding load intervals according to the real-time unit output value bound to each data point. The load interval is a continuous output segmentation based on the boiler's rated steam output. It can be divided into multiple standard intervals based on the general operating characteristics of industrial pulverized coal boilers, or the segmentation thresholds can be customized according to the rated parameters of a specific boiler. A unique load interval label is added to each time-series data point. Then, for each divided load interval, historical data that meets the baseline operating conditions within a recent period are statistically analyzed, i.e., data from boilers that have continuously and stably operated for a relatively long time within the same load interval, and NO... x The average set of historical operating data, including all ammonia slip compliance, no ash accumulation in the air preheater, no coking in the furnace, and unit steam coal consumption at the optimal level within the range, is used to calculate the emission benchmark, air preheater differential pressure benchmark, furnace outlet flue gas temperature trend benchmark, and coal consumption benchmark for that range. Then, all benchmark data are bound to the corresponding load range number and stored in the time series database. The benchmark values ​​are updated periodically to adapt to the boiler aging characteristics. Finally, when executing various diagnostic comparison logics, the load range label bound to the current data to be diagnosed is extracted, and only the benchmark operating conditions of the same load range or similar load ranges with small output differences are retrieved for parameter difference calculation. Cross-range comparisons are prohibited.

[0026] By dividing the load range into independent load intervals according to the preset output range, and establishing four exclusive benchmark operating conditions for each interval, the diagnostic process is forced to limit parameter comparison to only within the same or similar intervals, thus completely avoiding misjudgment problems caused by load differences and improving the accuracy of operating condition matching.

[0027] In this embodiment of the application, NO in S4 x Co-optimization diagnostics with ammonia slip include: S41, obtaining NO within the same load range x Emission concentration, ammonia slip concentration, ammonia injection valve opening, air volume, air temperature, and oxygen content; S42, based on NO x Emission status is determined by emission limits and ammonia slip limits; S43, Filter historical operating data to simultaneously meet the NO... x Candidate operating conditions for emission limits and ammonia slip limits; S44, Based on the candidate operating conditions, generate recommended parameters including air volume, air temperature, oxygen content, and ammonia injection valve opening; S45, When NO x When emissions or ammonia escape exceed limits, corresponding coordinated adjustment suggestions will be provided.

[0028] In some embodiments, the same load interval bound to the current data to be diagnosed is first locked, and the NO values ​​within that interval are read synchronously. x(Refers to nitrogen oxides, the most significant air pollutant in pulverized coal boiler flue gas, for which national and local environmental protection departments have set strict emission limits) emission concentration, ammonia slip (refers to free ammonia in the denitrification system that has not participated in the reduction reaction; when the concentration exceeds the limit, it will react with sulfur trioxide in the flue gas to form ammonium bisulfate, which corrodes the metal heat exchange fins of the air preheater and blocks the heat exchange channels) concentration, real-time opening of the denitrification ammonia injection regulating valve, total primary and secondary air volume, air preheater outlet combustion air temperature, and real-time oxygen content at the furnace outlet. These six core related parameters are then retrieved from the cloud-based preset plant environmental control limits. The real-time concentration values ​​are compared with the limits, and those meeting the limits are marked as meeting both standards or only NO. x The system identifies four emission states: exceeding limits, ammonia escape exceeding limits only, and simultaneous exceeding limits for both indicators. It then iterates through all historical time-series data stored for the recent period within that load range, automatically removing invalid periods such as unit start-ups / shutdowns and significant load fluctuations. It then filters out all stable operating periods that simultaneously meet both environmental limits, and summarizes these to generate multiple candidate operating conditions (i.e., simultaneously meeting NO emission limits within the same load range). x The dataset contains historical continuous operating data (emission limits, ammonia slip limits, and unit load fluctuations are relatively small, continuous operating time is relatively long, and unit steam coal consumption is at a moderate to optimal level within the range). Then, all candidate operating condition datasets are categorized by NO. x The emission concentration, ammonia slip concentration, and coal consumption level are weighted and comprehensively scored. The operating conditions with higher scores are selected and their averages are statistically analyzed. Standardized recommended parameters for this load range are output, including the recommended total air volume control range, the primary and secondary air ratio range, the optimal control range for furnace oxygen, and the optimal opening range for the ammonia injection regulating valve. Finally, when NO is detected... x When emissions or ammonia escape concentrations exceed limits, corresponding coordinated adjustment suggestions will be output based on the actual type of exceedance.

[0029] By achieving NO within the same load range x Synergistic analysis with ammonia slip as a dual indicator avoids alternating exceedances of the two pollutants due to single indicator adjustments. It automatically filters historical high-quality operating conditions that meet both standards to generate standardized recommended operating parameters, replacing the crude adjustments relying on manual experience. Combined with comprehensive judgment of multiple combustion-side parameters, it can also reduce NO at the source. x The production volume should balance environmental compliance with the economic efficiency of unit operation.

[0030] In this embodiment of the application, the coordinated adjustment suggestion in S45 includes: S451, when NO x When the emission concentration exceeds the limit but the ammonia slip concentration does not exceed the limit, priority should be given to generating enhanced adjustment suggestions for ammonia injection valve opening, air volume, or oxygen quantity; S452, when the ammonia slip concentration exceeds the limit and NO x When the emission concentration does not exceed the limit, suggestions for adjusting the ammonia injection valve opening or the air volume ratio should be generated first; S453, when NO xWhen both emission concentration and ammonia slip concentration exceed the limit, joint optimization suggestions for combustion-side parameters and denitrification-side parameters are generated.

[0031] In some embodiments, the current abnormal scenario is first determined, and when NO x When the emission concentration exceeds the limit but the ammonia escape concentration does not exceed the limit, the root cause of the fault is determined to be insufficient total ammonia injection or excessively high furnace oxygen levels exacerbating NO. x Upon generation, an enhanced adjustment suggestion is prioritized, initially slightly increasing the ammonia injection valve opening to replenish the denitrification reducing agent. If NO... x If the target is still not met, further increase the secondary air volume and lower the furnace oxygen content to the recommended range to reduce furnace NO from the source. x The generation rate is such that it is not recommended to lower the ammonia injection valve throughout the process to avoid ammonia escape exceeding the limit. Next, determine if the current abnormal scenario is that the ammonia escape concentration exceeds the limit and NO... x If the emission concentration does not exceed the limit, the root cause of the fault is determined to be excessive ammonia injection or uneven mixing of flue gas and ammonia. In this case, the ammonia injection valve opening should be reduced first to decrease the total ammonia injection. If ammonia escape still exceeds the limit after a single adjustment, the primary and secondary air ratio should be optimized to increase flue gas turbulence, improve denitrification reaction efficiency, and reduce residual ammonia escape. Finally, the current abnormal scenario is determined to be NO. x When both combustion and denitrification escape concentrations exceed the limit, it is determined to be a complex anomaly caused by the superposition of deterioration in combustion conditions and malfunction in the denitrification system. In this case, joint optimization suggestions for the combustion and denitrification sides are simultaneously output, firstly by fine-tuning the primary and secondary air ratio and furnace oxygen content to reduce furnace NO. x To reduce excessive ammonia injection, the total amount generated is adjusted by slightly reducing the opening of the ammonia injection valve. The parameters are then gradually adjusted step by step, referencing the recommended parameters for candidate operating conditions within the same load range. The changes in the two environmental protection indicators are monitored simultaneously until all meet the standards.

[0032] By establishing clear adjustment priorities for three types of abnormal scenarios, local fine-tuning is used to ensure stable unit operation when a single indicator exceeds the limit, while a joint optimization strategy on the combustion side and the denitrification side is adopted when two indicators simultaneously exceed the limit, starting from the furnace NO x Simultaneous regulation of both the source of emissions and the denitrification reaction shortens the time required to meet emission standards.

[0033] In this embodiment of the application, the air preheater ash accumulation classification diagnosis in S4 includes: S461, obtaining the negative pressure difference on side A and side B of each section of the air preheater, wherein the section includes at least one of the following: high temperature section on the primary air side, medium temperature section on the primary air side, low temperature section on the primary air side, high temperature section on the secondary air side, medium temperature section on the secondary air side, and low temperature section on the secondary air side; S462, within the same load range, calculating the deviation value of the negative pressure difference on side A relative to the reference negative pressure difference on side A and the deviation value of the negative pressure difference on side B relative to the reference negative pressure difference on side B, wherein the reference negative pressure difference on side A and the reference negative pressure difference on side B are derived from the air preheater pressure difference reference; S463, within the same load range, calculating... The A / B side difference value between the negative pressure difference on side A and the negative pressure difference on side B; S464. Based on the deviation value on side A, the deviation value on side B, and the A / B side difference value, classify the ash accumulation status into normal, light ash accumulation, moderate ash accumulation, and severe ash accumulation; S465. Output the soot blowing section, soot blowing priority, and suggested handling time based on the ash accumulation status; S466. When the same section reaches moderate or severe ash accumulation in multiple consecutive load intervals, increase the soot blowing priority of the corresponding section; S467. When multiple sections in the low-temperature section and the medium-temperature section of the primary air side, and the low-temperature section and the medium-temperature section of the secondary air side simultaneously reach moderate or severe ash accumulation, generate a combined soot blowing suggestion.

[0034] In some embodiments, firstly, real-time negative pressure difference data of the A and B sides of all six sections of the air preheater (i.e., the air preheater, a rotary heat exchanger that uses the waste heat of high-temperature flue gas from the boiler to heat the combustion air, divided into two independent heat exchange sides, A and B, with internal channels divided into primary air channels and secondary air channels, each channel further divided into three independent heat exchange areas: high-temperature section, medium-temperature section, and low-temperature section) are collected. Next, the load interval to which the current data belongs is locked, and the standard negative pressure difference of the A side and the standard negative pressure difference of the B side without ash accumulation stored in that interval are retrieved. The deviation values ​​are calculated separately. Then, the A / B side difference value of the real-time pressure difference of the A side and the real-time pressure difference of the B side in the same heat exchange section is calculated to determine whether unilateral local ash accumulation blockage has occurred. Finally, the three quantitative indicators of the A side deviation value, the B side deviation value, and the A / B side difference value are combined to classify the ash accumulation status into normal, light ash accumulation, and moderate ash accumulation. The system identifies four levels of ash accumulation: moderate, severe, and high. Based on the ash accumulation level of each section, a corresponding soot blowing strategy is output. Under normal conditions, no blowing is required. For light ash accumulation, the unit maintains its regular cycle blowing. For moderate ash accumulation, the fixed blowing cycle is shortened. For severe ash accumulation, an immediate blowing command is triggered. Simultaneously, the corresponding soot blowing section (the blowing area of ​​the independent steam soot blower corresponding to each high, medium, and low temperature section of the air preheater, each section equipped with an independent steam valve and soot blower) number and suggested single blowing duration are marked. When the same heat exchange section is determined to have moderate or severe ash accumulation in multiple consecutive load intervals, it is considered stubborn ash blockage, and the soot blowing priority of that section is directly adjusted to the highest, prioritizing the allocation of high-temperature blowing steam. Finally, when multiple sections in the primary air low-temperature section, primary air medium-temperature section, secondary air low-temperature section, and secondary air medium-temperature section are simultaneously determined to have moderate or severe ash accumulation, a multi-section combined soot blowing suggestion is automatically generated. This controls the multi-section soot blowers to blow continuously in sequence, avoiding excessively long blowing times and persistently low heat exchange efficiency for individual sections.

[0035] By collecting independent pressure differentials on sides A and B of all heat exchange sections of the air preheater, the fault points of local ash accumulation on one side and in one section can be accurately located. The ash accumulation status is classified by multiple dimensions, including the deviation of the pressure differential and the difference between the two sides, which improves the accuracy of the ash accumulation severity assessment. The soot blowing priority of each section is dynamically adjusted according to the ash accumulation level and the duration of the abnormality. When multiple sections have simultaneous heavy ash accumulation, a combined soot blowing scheme is automatically generated to reasonably allocate the purging steam and reduce the heat loss of flue gas.

[0036] In this embodiment, the prediction of coking trend of the heating surface in S4 includes: S471, extracting the time series of flue gas temperature at the furnace outlet within the same load range; S472, performing trend fitting on the time series to obtain the slope of the flue gas temperature at the furnace outlet as a function of operating time; S473, generating a coking warning for the heating surface when the slope exceeds a preset slope threshold or the cumulative increase in flue gas temperature at the furnace outlet exceeds a preset temperature rise threshold; S474, based on the slope or cumulative temperature rise of the flue gas temperature at the furnace outlet, and in conjunction with the boiler operating load, coal consumption, and NO... x The trend of emission concentration or ammonia escape concentration is used to determine the coking risk level; in particular, when the flue gas temperature at the furnace outlet continues to rise and coal consumption rises abnormally or NO... x When the emission concentration or ammonia escape concentration exceeds the limit, the coking risk level is increased.

[0037] In some embodiments, firstly, the load range bound to the current data to be diagnosed is locked, and the flue gas temperature data of the furnace outlet collected continuously over a period of time within this range is extracted to form a complete time series dataset, that is, a continuous time series array composed of multiple sets of flue gas temperature data of the furnace outlet collected continuously at a fixed sampling period within the same load range. Interference data with large load fluctuations due to unit start-up and shutdown are automatically removed. Then, the least squares method is used to perform linear fitting on the flue gas temperature time series dataset to calculate the trend fitting slope value of the flue gas temperature random group running time change, that is, the flue gas temperature per unit time obtained after linear fitting of the flue gas temperature time series using the fitting algorithm. The change in slope indicates a continuous unidirectional increase in flue gas temperature. Simultaneously, the cumulative increase in flue gas temperature at the furnace outlet during the specified period is calculated. Then, a pre-set dual warning threshold is retrieved from the cloud. If the fitted slope exceeds the slope threshold, or the cumulative temperature rise during the period exceeds the temperature rise threshold, a coking warning is immediately generated for the primary heating surfaces (including boiler furnace water-cooled walls, screen-type superheaters, reheaters, and other high-temperature metal heat exchange walls; ash from pulverized coal combustion melts and adheres to these walls at high temperatures, forming hard coke blocks that block flue gas flow channels and reduce heat exchange efficiency). Finally, a secondary correction logic for the coking risk level is executed, taking into account the unit's real-time operating load, instantaneous coal consumption, and NOx. x The risk level is adjusted based on four linked parameters: emission concentration, ammonia slip concentration, etc. If the following conditions are met simultaneously: continuous rise in flue gas temperature, abnormal increase in coal consumption, and NO... x If any ammonia escape indicator exceeds the limit, it is determined that the coke block is continuously thickening and is directly upgraded to the highest level of coking risk. If only the flue gas temperature rises and all other parameters are normal, it is determined to be a low-risk load fluctuation temperature rise, and only a prompt message is pushed, without triggering an emergency operation and maintenance warning.

[0038] By performing trend analysis based on continuous time series of flue gas temperature at the furnace outlet, rather than single-point instantaneous values, the slow-growing coking trend can be detected in advance. Early warnings can be triggered by setting dual thresholds for flue gas temperature change slope and cumulative temperature rise, reducing false alarms caused by short-term load fluctuations in the unit, and linking coal consumption and NOx levels. x The system uses multiple parameters to correct coking risk levels for ammonia escape, accurately distinguishing between normal load temperature rise and abnormal temperature rise due to coking blockage, and taking preventative measures to avoid furnace safety accidents.

[0039] In this embodiment, the coal consumption energy efficiency analysis in S4 includes: S481, calculating the unit steam coal consumption based on coal consumption and steam output; S482, calculating the operational stability index based on boiler operating load, main steam pressure, and main steam temperature; S483, comparing the unit steam coal consumption and operational stability index with historical or design benchmarks; S484, when the unit steam coal consumption abnormally increases, triggering NO2. x The results of co-optimization diagnosis of ammonia escape, graded diagnosis of air preheater ash accumulation, and prediction of coking trend of heated surfaces are used to generate an energy efficiency cause analysis.

[0040] In some embodiments, the instantaneous total coal consumption of the boiler and the real-time total steam output of the unit are first collected in real time to calculate the unit steam coal consumption. Then, the real-time load, main steam pressure, and main steam temperature of the unit are collected simultaneously, and the fluctuation range of the three parameters over a recent period is calculated. The parameters are then weighted and summed to obtain the operating stability index (a quantitative index obtained by weighting the fluctuation range of the unit load, main steam pressure, and main steam temperature; the larger the parameter fluctuation range, the lower the stability score). Then, the historical optimal coal consumption benchmark and the boiler's factory rated design benchmark for the current load range are retrieved, i.e., the theoretical optimal unit steam coal consumption and optimal operating parameters under rated operating conditions provided by the manufacturer when the boiler leaves the factory. These are the standard reference values ​​for energy efficiency evaluation. The real-time unit steam coal consumption and operating stability index are compared with the two benchmarks, and marked as three states: normal energy efficiency, slightly abnormal energy efficiency, and severely abnormal energy efficiency. Finally, when the unit steam coal consumption is higher than the interval benchmark threshold and energy efficiency is judged to be abnormal, the NO from the same period is automatically retrieved. x All diagnostic data from ammonia escape co-diagnosis, air preheater ash accumulation classification diagnosis, and heating surface coking trend prediction are automatically used to generate an energy efficiency root cause analysis. If ash accumulation or coking is determined to be moderate or above abnormal, the increased coal consumption is due to heat exchange channel blockage and increased flue gas heat loss. If the denitrification dual indicators frequently exceed limits, the cause is unreasonable combustion parameter control and furnace NO2. x If the generation rate is too high or there are abnormalities in the synchronization of multiple modules, a comprehensive analysis of the increased energy consumption and targeted optimization suggestions will be provided.

[0041] By quantifying unit steam coal consumption and operational stability indicators, and benchmarking against historical and design benchmarks, energy efficiency anomalies can be quickly identified. When coal consumption rises abnormally, the other three diagnostic modules are automatically linked to comprehensively analyze the causes, accurately locate the root cause of the energy consumption increase, and guide operation and maintenance personnel to optimize equipment or combustion parameters in a targeted manner to improve the economic efficiency of boiler operation.

[0042] Example 2:

[0043] Reference Figure 2 This application also provides a remote intelligent diagnostic system for pulverized coal boilers based on a one-way secure acquisition link, comprising: a field data acquisition layer for acquiring boiler operation data from the DCS of the pulverized coal boiler via OPC communication service; a one-way secure transmission layer connected to the field data acquisition layer for encapsulating the boiler operation data into UDP packets by the field acquisition service, enabling the UDP packets to be transmitted unidirectionally in the forward direction via a one-way isolation gateway, and having the UDP packets converted into MQTT data topics and uploaded to the cloud platform by an external network-side forwarding service; a cloud preprocessing layer connected to the one-way secure transmission layer for performing data cleaning, time alignment, outlier processing, measurement point mapping, and load range identification on the boiler operation data on the cloud platform to form a dataset to be diagnosed; and a multi-module diagnostic layer connected to the cloud preprocessing layer for performing NO on the dataset to be diagnosed. x The system integrates ammonia escape optimization diagnosis, air preheater ash accumulation classification diagnosis, heating surface coking trend prediction, and coal consumption energy efficiency analysis; the diagnostic fusion output layer, connected to the multi-module diagnostic layer, is used to fuse the diagnostic results to generate early warning levels, handling priorities, operation optimization suggestions, and periodic diagnostic reports.

[0044] In some embodiments, the field sensing layer (including a combination of hardware and software deployed in the boiler's on-site industrial control room, with the hardware being an industrial control computer and the software being an OPC standard communication service and data encapsulation program, possessing only read-only access to DCS data) first continuously connects to the pulverized coal boiler's DCS distributed control system through the OPC communication service deployed on the industrial control computer. It reads all raw time-series data from all operating measurement points in real time and caches short-term measurement point data locally. Then, the network transmission layer (i.e., the intermediate layer connecting the field and the cloud, with hardware including a one-way isolation gateway and software including a field acquisition and encapsulation service and an external network-side MQTT forwarding gateway service) is wired to the field sensing layer's industrial control computer via a network cable. The field acquisition service encapsulates the raw measurement point data into UDP packets with a unified timestamp and sends them to the one-way isolation gateway to complete the forward one-way transmission from the industrial control network to the external network. The external network-side forwarding gateway receives the UDP packets, converts them into standardized MQTT topic messages, and transmits them to the cloud server cluster via the public network. Finally, the cloud platform... The core computing and storage layer (deployed on a cloud server, including a time-series database, intelligent diagnostic engine, report generation module, and API service) establishes a long connection with the MQTT gateway in the network transport layer to continuously receive messages. It writes the data into the time-series database for persistent storage and sequentially performs data cleaning, timestamp alignment, outlier removal, unified mapping of measurement point codes, and load interval label binding preprocessing. It outputs a structured and standardized dataset to be diagnosed. Then, the intelligent diagnostic engine in the cloud platform layer starts four independent diagnostic operation programs in parallel to simultaneously complete denitrification synergistic optimization, air preheater ash grading, coking trend prediction, and coal consumption and energy efficiency analysis. It independently stores the original diagnostic result data packets of each module. Finally, the application display layer (deployed in the cloud as a human-computer interaction layer, including Web visualization service, mobile early warning push service, and automatic report generation service) connects with the cloud platform layer to read the integrated diagnostic results and uniformly visualize and push real-time monitoring data, early warning information, optimization suggestions, and periodic diagnostic reports.

[0045] In this embodiment, the multi-module diagnostic layer includes: an emission synergistic optimization module, used to perform the NO... x The system includes a collaborative optimization diagnostic module for ammonia slip; an ash accumulation diagnostic module for performing ash accumulation classification diagnostics on the air preheater; a coking prediction module for predicting the coking trend of the heating surface; and an energy efficiency analysis module for performing coal consumption energy efficiency analysis.

[0046] In some embodiments, the time-series database first stores all boiler operation data uploaded by the network transport layer, providing data support for all upper-layer modules. Then, the diagnostic engine reads the standardized dataset to be diagnosed from the time-series database, in which the emission co-optimization module independently executes NO... xThe system performs all calculations for the collaborative optimization diagnosis of ammonia slip, and independently performs all calculations for the ash accumulation diagnosis of the air preheater. The coking prediction module independently performs all calculations for the prediction of coking trends on the heating surface, and the energy efficiency analysis module independently performs all calculations for the analysis of coal consumption and energy efficiency. Then, the diagnosis result fusion module receives the original diagnosis results from the four sub-modules, completes the mapping of early warning levels, the calculation of disposal priorities, and the integration of optimization suggestions. Subsequently, the report generation module retrieves corresponding data from the time series database and the diagnosis engine according to the preset template and automatically generates periodic diagnosis reports. Finally, the API service encapsulates all data and results of the platform into standardized interfaces to provide data query and call services to the outside world.

[0047] In this embodiment of the application, the one-way isolation gateway in the one-way secure transmission layer is configured to only allow forward data transmission from the industrial control network side to the external network side, and the external network-side forwarding service in the one-way secure transmission layer is configured to generate MQTT topics or message fields according to the device number, boiler number, measurement point type and timestamp.

[0048] In some embodiments, the unidirectional isolation gateway is first configured for hardware transmission. The underlying physical transmission link of the hardware only opens a unidirectional data transmission channel from the industrial control intranet to the external network. The hardware chip completely cuts off the physical transmission path from the external network to the industrial control intranet, and there is no reverse instruction or reverse measurement point data return channel. A built-in dedicated data packet filtering chip automatically intercepts all data packets with the external network as the source address and the boiler DCS industrial control device as the destination address. Only the data packets of the unit operation measurement points collected by the DCS are allowed to flow unidirectionally to the external cloud. Then, the MQTT topic is configured for the external network forwarding service. The MQTT topic is generated using a four-layer fixed layer splicing format. The layer fields are, in order, the acquisition gateway device number, the boiler unit number, the measurement point type code, and the precise timestamp, ensuring that each message can uniquely identify its source device, boiler, measurement point, and acquisition time. Finally, the cloud time series database stores the data in partitions according to the layer information of the MQTT topic, based on the device, boiler, and measurement point type. When querying historical data, it can be directly filtered by boiler number and measurement point code, which greatly improves data retrieval efficiency and cloud computing speed.

[0049] Example 3:

[0050] The field perception layer is deployed in the boiler control room, including one industrial control computer and one DCS system interface card. The industrial control computer is connected to the boiler DCS system via Ethernet and is equipped with OPC communication service and field data acquisition service software. The network transmission layer is deployed in the field computer room and external network access area, including one one-way isolation gateway, one 4G / 5G router, and one external network front-end server. The internal network port of the one-way isolation gateway is connected to the field industrial control computer, and the external network port is connected to the external network front-end server. The external network front-end server is equipped with MQTT forwarding service software and is connected to the public network via the 4G / 5G router. The cloud platform layer is deployed on a public cloud server cluster, including a time-series database server, an intelligent diagnostic computing server, a web application server, and an API server. The servers are connected to each other via a high-speed internal network. The application display layer supports access via web browser and mobile APP. Maintenance personnel can view boiler operation data, receive early warning information, and download diagnostic reports anytime, anywhere via computer or mobile phone. The entire network architecture adopts a layered isolation design. The industrial control network and the external network are physically isolated by a one-way isolation gateway, and encrypted transmission is used between the external network and the cloud, which comprehensively protects the security of data transmission and system operation.

[0051] The system supports automatic generation of periodic diagnostic reports in PDF format on a daily, weekly, and monthly basis. The reports comprehensively cover all dimensions of boiler operation information, including: basic project information (boiler model, serial number, operating time, and reporting cycle); a recent cycle operation overview (summarizing overall boiler operational stability, load distribution, and average values ​​of key operating parameters); coal consumption analysis (showing the trend of unit steam coal consumption, comparison with historical and design benchmarks, and analysis of periods of abnormal energy efficiency); NOx and ammonia escape analysis (statistics on environmental compliance rates, periods and causes of exceedances, and recommended adjustments to operating parameters); air preheater ash accumulation analysis (showing changes in ash accumulation in each section, soot blowing records, and soot blowing optimization suggestions); heating surface coking analysis (presenting furnace outlet flue gas temperature trends, coking risk levels, and early warning handling status); an early warning list (summarizing all early warning information, warning levels, and handling status within the cycle); optimization suggestions (providing targeted operational adjustments and equipment maintenance recommendations based on the full-cycle diagnostic results); and historical trend charts (visually displaying load, coal consumption, NOx, and other parameters). x The changing trends of key parameters such as ammonia slip and air preheater differential pressure.

[0052] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

[0053] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not describe the various possible combinations separately.

[0054] Furthermore, various different implementations of the present invention can be combined arbitrarily, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed in the present invention.

Claims

1. A remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure data acquisition link, characterized in that, Includes the following steps: S1. Obtain boiler operation data from the DCS of the pulverized coal boiler through OPC communication service; S2. The on-site data acquisition service encapsulates the boiler operation data into UDP packets, and the UDP packets are transmitted in a forward unidirectional manner through a one-way isolation gateway. The external network forwarding service then converts the UDP packets into MQTT data topics and uploads them to the cloud platform. S3. On the cloud platform, the boiler operation data is cleaned, time aligned, outlier processed, measurement point mapped, and load range identified to form a dataset to be diagnosed. S4. Execute NO based on the dataset to be diagnosed. x Collaborative optimization diagnosis with ammonia slip, graded diagnosis of air preheater ash accumulation, prediction of coking trend of heating surface, and analysis of coal consumption efficiency; S5. Integrate the diagnostic results to generate early warning levels, handling priorities, operational optimization suggestions, and periodic diagnostic reports.

2. The remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link according to claim 1, characterized in that, The load interval identification in S3 includes: S31. Divide the continuous operation data into multiple load intervals according to the preset load range; S32. Establish emission benchmarks, air preheater differential pressure benchmarks, furnace outlet flue gas temperature trend benchmarks, and coal consumption benchmarks for each load range; S33. During diagnosis, only the current operating condition is compared with the baseline operating condition within the same or similar load range.

3. The remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link according to claim 1, characterized in that, NO in S4 x Co-optimized diagnosis with ammonia escape includes: S41. Obtain NO within the same load range x Emission concentration, ammonia slip concentration, ammonia injection valve opening, air volume, air temperature, and oxygen content; S42, According to NO x Emission limits and ammonia slip limits are used to determine emission status; S43. Filter from historical operating data that simultaneously meet the above NO x Candidate operating conditions for emission limits and ammonia slip limits; S44. Generate recommended parameters, including air volume, air temperature, oxygen content, and ammonia injection valve opening, based on candidate operating conditions; S45, When NO x When emissions or ammonia escape exceed limits, corresponding coordinated adjustment suggestions will be provided.

4. The remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link according to claim 3, characterized in that, The coordinated adjustment recommendations in S45 include: S451, When NO x When the emission concentration exceeds the limit but the ammonia slip concentration does not exceed the limit, priority should be given to generating enhanced adjustment suggestions for the ammonia injection valve opening, air volume, or oxygen quantity. S452, When the ammonia escape concentration exceeds the limit and NO x When the emission concentration does not exceed the limit, suggestions for reducing the opening of the ammonia injection valve or adjusting the air volume ratio will be generated first. S453, When NO x When both emission concentration and ammonia slip concentration exceed the limit, joint optimization suggestions for combustion-side parameters and denitrification-side parameters are generated.

5. The remote intelligent diagnostic method for pulverized coal boilers based on a unidirectional secure acquisition link according to claim 1, characterized in that, The air preheater ash accumulation classification diagnosis in S4 includes: S461. Obtain the negative pressure difference on side A and side B of each section of the air preheater, wherein the section includes at least one of the following: high temperature section on the primary air side, medium temperature section on the primary air side, low temperature section on the primary air side, high temperature section on the secondary air side, medium temperature section on the secondary air side, and low temperature section on the secondary air side. S462. Within the same load range, calculate the deviation of the negative pressure difference on side A from the reference negative pressure difference on side A and the deviation of the negative pressure difference on side B from the reference negative pressure difference on side B, wherein the reference negative pressure difference on side A and the reference negative pressure difference on side B are derived from the air preheater pressure difference reference. S463. Within the same load range, calculate the A / B side difference value between the negative pressure difference on side A and the negative pressure difference on side B; S464. Based on the deviation value of side A, the deviation value of side B, and the difference value of side A / B, the dust accumulation status is divided into normal, light dust accumulation, moderate dust accumulation, and severe dust accumulation. S465. Output the soot blowing section, soot blowing priority, and suggested disposal time according to the ash accumulation status; S466. When the same section reaches moderate or severe dust accumulation in multiple consecutive load intervals, the soot blowing priority of the corresponding section shall be increased. S467. When multiple sections of the low-temperature section and the medium-temperature section of the primary air side, and the low-temperature section and the medium-temperature section of the secondary air side simultaneously reach moderate or severe dust accumulation, a combined soot blowing suggestion is generated.

6. The remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link according to claim 1, characterized in that, The prediction of coking trend on the heated surface in S4 includes: S471. Extract the time series of flue gas temperature at the furnace outlet within the same load range; S472. Perform trend fitting on the time series to obtain the slope of the furnace outlet flue gas temperature as a function of operating time; S473. When the slope exceeds the preset slope threshold or the cumulative increase in flue gas temperature at the furnace outlet exceeds the preset temperature rise threshold, a coking warning for the heated surface is generated. S474. Based on the slope or cumulative temperature rise of the flue gas temperature at the furnace outlet, and in conjunction with the boiler operating load, coal consumption, and NO... x The trend of emission concentration or ammonia slip concentration is used to determine the coking risk level; in particular, when the flue gas temperature at the furnace outlet continues to rise and coal consumption rises abnormally or NO... x When the emission concentration or ammonia escape concentration exceeds the limit, the coking risk level is increased.

7. The remote intelligent diagnostic method for pulverized coal boilers based on a one-way secure acquisition link according to claim 1, characterized in that, The coal consumption energy efficiency analysis in S4 includes: S481. Calculate the unit steam coal consumption based on coal consumption and steam production. S482. Calculate the operational stability index based on the boiler operating load, main steam pressure, and main steam temperature. S483. Compare the unit steam coal consumption and operational stability indicators with historical or design benchmarks; S484. When the unit steam coal consumption increases abnormally, the NOx will be activated. x The results of co-optimization diagnosis of ammonia slip, graded diagnosis of air preheater ash accumulation, and prediction of coking trend of heated surfaces are used to generate an energy efficiency cause analysis.

8. A remote intelligent diagnostic system for pulverized coal boilers based on a one-way secure data acquisition link, characterized in that: The method for remote intelligent diagnosis of pulverized coal boilers based on a one-way secure acquisition link, as described in any one of claims 1-7, includes: The field data acquisition layer is used to acquire boiler operation data from the DCS of the pulverized coal boiler via OPC communication service; A one-way secure transmission layer, connected to the field data acquisition layer, is used to encapsulate the boiler operation data into UDP packets by the field acquisition service, so that the UDP packets are transmitted in a forward one-way direction through the one-way isolation gateway, and the external network forwarding service converts the UDP packets into MQTT data topics and uploads them to the cloud platform. The cloud preprocessing layer, connected to the one-way secure transmission layer, is used to perform data cleaning, time alignment, outlier processing, measurement point mapping, and load range identification on the boiler operation data on the cloud platform to form a dataset to be diagnosed. A multi-module diagnostic layer, connected to the cloud preprocessing layer, is used to perform NO tests based on the dataset to be diagnosed. x Collaborative optimization diagnosis with ammonia slip, graded diagnosis of air preheater ash accumulation, prediction of coking trend of heating surface, and analysis of coal consumption efficiency; The diagnostic fusion output layer, connected to the multi-module diagnostic layer, is used to fuse the diagnostic results and generate early warning levels, handling priorities, operational optimization suggestions, and periodic diagnostic reports.

9. The remote intelligent diagnostic system for pulverized coal boilers based on a one-way secure acquisition link according to claim 8, characterized in that, The multi-module diagnostic layer includes: The emission co-optimization module is used to execute the NO x Synergistic optimization of diagnosis with ammonia escape; The ash accumulation diagnosis module is used to perform ash accumulation classification diagnosis of the air preheater; The coking prediction module is used to perform coking trend prediction on the heated surface; The energy efficiency analysis module is used to perform the coal consumption energy efficiency analysis.

10. The remote intelligent diagnostic system for pulverized coal boilers based on a unidirectional secure acquisition link according to claim 8, characterized in that, The one-way isolation gateway in the one-way secure transmission layer is configured to only allow forward data transmission from the industrial control network side to the external network side. The external network-side forwarding service in the one-way secure transmission layer is configured to generate MQTT topics or message fields according to the device number, boiler number, measurement point type, and timestamp.

Images