Coal-fired unit deep peak shaving coal mixed AI collaborative decision and fault early warning method

Abstract

The present application relates to a kind of coal-fired mixed AI collaborative decision-making and fault early warning method for deep peak shaving of coal-fired unit, to solve the problem that fuel deployment and equipment state early warning are disconnected in the prior art, rely on fixed threshold alarm and static coal blending strategy, which leads to high risk of operation of unit under deep peak shaving condition, poor adaptability, by generating multiple initial coal blending schemes according to target load and unit state, and real-time monitoring of operating parameters to trigger dynamic adjustment. The scheme optimization process considers stability, economy and equipment adaptability, selects the optimal scheme, establishes a safety operation interval that changes with load and coal quality, accurately monitors the state of key equipment, and implements graded warning and interlock operation suggestion, realizing the whole chain cooperation from fuel inlet to equipment state. The present application can significantly improve the safety and stability of coal-fired unit under deep peak shaving condition, and provide effective support for flexible and low-carbon operation of power plant.

Description

Technical Field

[0001] This application relates to the field of coal-fired unit condition monitoring technology, and in particular to a coal-fired unit deep peak shaving method for coal-fired hybrid AI collaborative decision-making and fault early warning. Background Technology

[0002] Against the backdrop of the current energy structure transformation, coal-fired power generating units are shifting from the traditional base load operation mode to undertaking grid peak shaving tasks, and frequent participation in deep peak shaving has become the norm. Deep peak shaving means that the unit needs to operate under conditions far below the rated load, which leads to a significant reduction in boiler combustion stability and a sharp increase in the difficulty of steam parameter control, posing a severe challenge to the safe, environmental, and economical operation of the unit.

[0003] To address this challenge, existing technologies typically employ coal blending to adapt to low-load conditions. However, current coal blending methods largely rely on operator experience or simple static calorific value calculations, lacking a precise understanding of the dynamic coupling relationship between the characteristics of multiple coal types and complex operating conditions. Furthermore, during deep peak shaving, critical auxiliary equipment operates under off-design conditions, and its condition monitoring generally relies on fixed threshold alarms, failing to detect the impact of load and coal quality changes on equipment safety boundaries and hindering the timely detection of potential faults.

[0004] Furthermore, existing operation optimization and equipment early warning systems are often independent, forming information silos. Fuel allocation decisions fail to fully consider their cascading impact on the operating status of subsequent auxiliary equipment, and equipment early warnings fail to provide feedback to the fuel decision-making end to trigger pre-emptive adjustments. This disconnect between decision-making and early warning makes it difficult for units to achieve global safety and economic synergy when dealing with rapidly changing peak-shaving demands, lacking an intelligent collaborative decision-making and early warning mechanism that can span the entire chain from fuel inlet to equipment status. Summary of the Invention

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] According to a first aspect of the present invention, the present invention claims protection for a coal-fired power unit deep peak shaving method for coal-fired hybrid AI collaborative decision-making and fault early warning, comprising the following steps:

[0007] S1. Based on the target load command and the current status of the unit, generate multiple initial coal mixing schemes according to a pre-set database of various typical coal quality characteristics.

[0008] S2. Real-time monitoring of the unit's operating parameters under deep peak shaving conditions, including boiler main steam pressure, main steam temperature, reheat steam temperature, furnace negative pressure, flue gas oxygen content, and pulverizing system parameters.

[0009] S3. Compare the operating parameters with the expected parameter range of the currently executed coal-fired power generation scheme. If the operating parameters continue to deviate from the expected parameter range, trigger a dynamic adjustment command for the power generation scheme.

[0010] S4. In response to the dynamic adjustment command of the mixing scheme, the multiple initial coal-fired mixing schemes are screened and sorted based on real-time operating parameters, and the optimal coal-fired mixing scheme and the sequence of backup schemes are output.

[0011] S5. Execute the optimal coal mixing scheme and continuously monitor the status parameters of key equipment, including the temperature of the coal mill bearing, the current of the primary air fan, and the feedback of the coal feeder speed.

[0012] S6. Establish the dynamic safe operating range of the key equipment status parameters under the deep peak shaving condition. The upper and lower limits of the dynamic safe operating range are dynamically adjusted according to the changes in the random group load and coal-fired mixing scheme.

[0013] S7. Real-time comparison of the status parameters of the key equipment with their corresponding dynamic safe operating range. When any parameter exceeds its dynamic safe operating range and continues for a first preset time, a first-level warning signal is generated.

[0014] S8. After a Level 1 warning signal is generated, if the parameter further deteriorates to the point of exceeding the preset emergency threshold outside its dynamic safe operating range, a Level 2 fault alarm signal is generated, and the corresponding equipment maintenance or system operation instructions are recommended or executed.

[0015] Furthermore, step S1 also includes:

[0016] S1.1 Determine the target heat load requirement of the unit according to the target load command;

[0017] S1.2. Based on the target heat load demand and the available coal types and stock of the current unit, select at least two available basic coal types from the typical coal quality characteristic database;

[0018] S1.3. With the optimization objective of meeting the target heat load requirement and minimizing the total cost, calculate the theoretical values ​​of the mixed coal quality of each basic coal type under different proportions, and generate the multiple initial coal mixing schemes and their corresponding expected parameter ranges.

[0019] Furthermore, the step S4 of screening and sorting the multiple initial coal-fired mixing schemes based on real-time operating parameters specifically includes:

[0020] S4.1 Obtain the real-time operating parameters monitored in step S2, and identify the most significant operating deviation item at present;

[0021] S4.2 Based on the most significant operational deviation item, eliminate from the plurality of initial coal-fired mixing schemes any schemes that are expected to exacerbate the deviation.

[0022] S4.3 For the remaining candidate solutions, construct a multi-objective evaluation function, which comprehensively considers at least the stability improvement potential of the solution, the increase in execution cost, and the correlation impact on other operating parameters;

[0023] S4.4 Calculate the evaluation function score of each candidate scheme, sort them from high to low according to the evaluation function score, output the scheme with the highest score as the optimal coal-fired mixing scheme, and use the subsequent schemes as the backup scheme sequence.

[0024] Furthermore, the step S6 of establishing the dynamic safe operating range of the key equipment status parameters under the deep peak shaving condition specifically includes:

[0025] S6.1 For each type of critical equipment, a standard safe operating range for the critical equipment under rated load is preset;

[0026] S6.2 Establish a load-safety interval correction coefficient mapping table and define the correction coefficients for the upper and lower limits of the standard safe operating interval of each device under different load levels;

[0027] S6.3 Establish a coal quality characteristic-safety range influence factor mapping table to define the direction and magnitude of the influence of different key coal quality characteristic indicators on the standard safe operating range of specific equipment;

[0028] S6.4. Based on the current actual load of the unit, query the load-safety interval correction coefficient mapping table and make the first correction to the standard safe operating interval;

[0029] S6.5. Based on the key coal quality characteristic indicators of the currently implemented or soon-to-be-implemented coal-fired mixed combustion scheme, query the coal quality characteristic-safety range influence factor mapping table, and perform a second correction on the safe operating range after the first correction to obtain the dynamic safe operating range under the current operating conditions.

[0030] Furthermore, in step S7, the first preset duration is set differently according to different key equipment status parameters.

[0031] Furthermore, in step S7, when a Level 1 warning signal is generated, a list of one or more suspected causes that lead to the abnormality of the parameter is output synchronously.

[0032] Furthermore, the interlocking recommendation or execution of corresponding equipment maintenance or system operation instructions in step S8 includes:

[0033] It is recommended to switch to a coal-fired mixed scheme in the backup plan sequence, reduce the output of the corresponding coal mill, or increase the frequency of inspections of specific equipment.

[0034] Furthermore, the key coal quality characteristics mentioned in step S6.3 include, but are not limited to, volatile matter, ash content, moisture content, and ash fusion point on an as-received basis.

[0035] Furthermore, the pulverizing system parameters mentioned in step S2 include the coal mill outlet temperature, coal mill differential pressure, and primary air volume.

[0036] Furthermore, the method also includes step S9: recording each scheme adjustment instruction, early warning signal, alarm signal, and subsequent operation and system response, and forming a case library for periodic optimization and updating of the typical coal quality characteristic database, scheme generation rules, and dynamic safe operation range calculation rules.

[0037] This invention relates to an AI-driven collaborative decision-making and fault early warning method for coal-fired power units in deep peak shaving. It aims to address the problems of high operational risk and poor adaptability of units under deep peak shaving conditions caused by the disconnect between fuel allocation and equipment status early warning, reliance on fixed threshold alarms, and static coal blending strategies in existing technologies. The invention generates multiple initial coal-fired mixing schemes based on the target load and unit status, and monitors operating parameters in real time to trigger dynamic adjustments. The scheme optimization process comprehensively considers stability, economy, and equipment adaptability, selecting the optimal scheme. By establishing a safe operating range that dynamically changes with load and coal quality, it accurately monitors the status of key equipment and implements tiered early warning and interlocking operation suggestions, achieving full-chain coordination from fuel inlet to equipment status. This invention can significantly improve the safety and stability of coal-fired power units under deep peak shaving conditions, providing effective support for the flexible and low-carbon operation of power plants. Attached Figure Description

[0038] Figure 1 The flowchart illustrates the process of a coal-fired power unit deep peak shaving AI collaborative decision-making and fault early warning method for coal-fired power units, as claimed in this embodiment of the invention.

[0039] Figure 2 The flowchart of step S1 of the coal-fired power plant deep peak shaving AI collaborative decision-making and fault early warning method is shown in the embodiment of the present invention.

[0040] Figure 3 The flowchart of step S4 of the coal-fired power plant deep peak shaving AI collaborative decision-making and fault early warning method is shown in the embodiment of the present invention.

[0041] Figure 4The flowchart shows step S6 of the coal-fired power plant deep peak shaving AI collaborative decision-making and fault early warning method for coal-fired power plants, as claimed in this embodiment of the invention. Detailed Implementation

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0043] The terms "first," "second," and "third" in this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this application are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0044] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0045] According to the first embodiment of the present invention, referring to Figure 1 This invention claims protection for a coal-fired power unit deep peak shaving AI collaborative decision-making and fault early warning method for coal-fired power plant, comprising the following steps:

[0046] S1. Based on the target load command and the current status of the unit, generate multiple initial coal mixing schemes according to a pre-set database of various typical coal quality characteristics.

[0047] S2. Real-time monitoring of the unit's operating parameters under deep peak shaving conditions, including boiler main steam pressure, main steam temperature, reheat steam temperature, furnace negative pressure, flue gas oxygen content, and pulverizing system parameters.

[0048] S3. Compare the operating parameters with the expected parameter range of the currently executed coal-fired power generation scheme. If the operating parameters continue to deviate from the expected parameter range, trigger a dynamic adjustment command for the power generation scheme.

[0049] S4. In response to the dynamic adjustment command of the mixing scheme, the multiple initial coal-fired mixing schemes are screened and sorted based on real-time operating parameters, and the optimal coal-fired mixing scheme and the sequence of backup schemes are output.

[0050] S5. Execute the optimal coal mixing scheme and continuously monitor the status parameters of key equipment, including the temperature of the coal mill bearing, the current of the primary air fan, and the feedback of the coal feeder speed.

[0051] S6. Establish the dynamic safe operating range of the key equipment status parameters under the deep peak shaving condition. The upper and lower limits of the dynamic safe operating range are dynamically adjusted according to the changes in the random group load and coal-fired mixing scheme.

[0052] S7. Real-time comparison of the status parameters of the key equipment with their corresponding dynamic safe operating range. When any parameter exceeds its dynamic safe operating range and continues for a first preset time, a first-level warning signal is generated.

[0053] S8. After a Level 1 warning signal is generated, if the parameter further deteriorates to the point of exceeding the preset emergency threshold outside its dynamic safe operating range, a Level 2 fault alarm signal is generated, and the corresponding equipment maintenance or system operation instructions are recommended or executed.

[0054] In this embodiment, each step further includes:

[0055] The system receives target load instructions from the power grid dispatch system; acquires the current operating status of the generating units, including current load, coal level in each coal bunker, and operating status of each coal feeder; based on the target load instructions and the current status of the generating units, it calls up a pre-set database of various typical coal quality characteristics, which stores industrial analysis, elemental analysis, and combustion characteristic parameters of all possible coal types in the plant; based on the above information, with the core objective of meeting the total calorific value requirement under the target load and minimizing the total fuel cost, it calculates the feasibility of different coal type combinations and ratios, generates multiple initial coal mixing schemes with differences in key indicators such as calorific value, volatile matter, and sulfur content, and sets the expected range of key operating parameters for each scheme;

[0056] The unit's distributed control system collects operating parameters in real time under deep peak shaving conditions. These operating parameters include the main steam pressure, main steam temperature, reheat steam temperature, furnace negative pressure, and flue gas oxygen content on the boiler side, as well as the coal mill outlet temperature, coal mill differential pressure, and primary air volume on the pulverizing system side.

[0057] The monitored real-time operating parameters are compared one by one with the expected parameter range corresponding to the currently executed coal-fired hybrid scheme. If one or more operating parameters are found to continuously deviate from their expected parameter range, and the deviation exceeds the preset dead zone threshold and the duration reaches a preset judgment period, it is determined that the current operating condition does not match the expectations, and the hybrid scheme dynamic adjustment command is triggered.

[0058] In response to the triggered adjustment command, the system locks the most critical operational deviation parameter. Based on the properties of this deviation parameter, it re-evaluates multiple initial coal-fired mixing schemes and selects those that can theoretically effectively correct or mitigate the current major operational deviation. Then, it comprehensively evaluates these candidate schemes, taking into account factors such as the execution complexity of scheme switching, the current inventory of the required coal type, and the potential impact risk on other operational parameters that have not experienced deviations. Finally, it outputs the coal-fired mixing scheme with the optimal comprehensive evaluation result and generates a sequence containing 2-3 alternative schemes.

[0059] While executing the optimal coal-fired power mixing scheme, strengthen the condition monitoring of key power equipment; key equipment condition parameters include the motor bearing temperature and vibration value of the coal mill, the motor current and bearing temperature of the primary air fan, and the motor speed feedback value and casing temperature of the coal feeder.

[0060] For each key equipment status parameter monitored, a dynamic safe operating range is established that varies with operating conditions. This range is first set based on the safety limits provided by the equipment manufacturer under rated operating conditions. Then, based on the unit's current real-time load value, a preset "load-safety factor" lookup table is consulted to perform a first scaling correction on the upper and lower boundaries of the safety limits. Next, based on the dominant coal quality characteristics in the current coal-fired mixing scheme, another preset "coal quality-influence factor" lookup table is consulted to perform a second offset correction on the boundaries after load correction, ultimately obtaining a dynamic safe operating range applicable to the current specific operating conditions.

[0061] The system compares the monitored equipment status parameters with the calculated dynamic safe operating range in real time. When the actual value of a parameter exceeds the upper or lower limit of its dynamic safe operating range, and this exceeding state continues for a preset first time window, a first-level warning signal is generated for the equipment. This signal includes the equipment name, the abnormal parameter and its current value.

[0062] After generating a Level 1 warning signal, the abnormal parameter continues to be monitored. If the parameter value deteriorates further, not only continuously exceeding the dynamic safe operating range but also reaching the emergency action threshold calculated based on the range boundary, a Level 2 fault alarm signal is immediately generated. At the same time, the system automatically interlocks and triggers a predefined set of operation suggestions, which points to the most direct operation to handle the potential fault of the equipment. For example, it may suggest switching the coal-fired mixing scheme that is strongly related to the equipment to the first option in the sequence of alternative schemes, or suggest that the operator immediately reduce the output setting value of the equipment.

[0063] Furthermore, referring to Figure 2 Step S1 also includes:

[0064] S1.1 Determine the target heat load requirement of the unit according to the target load command;

[0065] S1.2. Based on the target heat load demand and the available coal types and stock of the current unit, select at least two available basic coal types from the typical coal quality characteristic database;

[0066] S1.3. With the optimization objective of meeting the target heat load requirement and minimizing the total cost, calculate the theoretical values ​​of the mixed coal quality of each basic coal type under different proportions, and generate the multiple initial coal mixing schemes and their corresponding expected parameter ranges.

[0067] In this embodiment, each step further includes:

[0068] Parse the target load command and convert it into the total heat load requirement required by the boiler;

[0069] Verify the real-time coal level data of each raw coal bunker of the unit to determine the list of currently available coal types and their available quantities;

[0070] With total heat load demand as a constraint and the lowest total cost of coal fed into the furnace as the optimization direction, two or three types of coal are selected from the list of available coal types for combination ratio calculation. During the calculation, the key indicators of the blended coal, such as the received basis lower heating value, received basis volatile matter, and received basis sulfur content, are estimated by querying the database of typical coal quality characteristics and using the weighted average method for each ratio.

[0071] Exclude those blending schemes whose key indicators after mixing exceed the allowable range of the unit design, and set the expected operating parameter range for all remaining feasible schemes, thereby generating multiple initial coal-fired blending schemes.

[0072] Furthermore, referring to Figure 3 Step S4, which involves screening and sorting the multiple initial coal-fired mixing schemes based on real-time operating parameters, specifically includes:

[0073] S4.1 Obtain the real-time operating parameters monitored in step S2, and identify the most significant operating deviation item at present;

[0074] S4.2 Based on the most significant operational deviation item, eliminate from the plurality of initial coal-fired mixing schemes any schemes that are expected to exacerbate the deviation.

[0075] S4.3 For the remaining candidate solutions, construct a multi-objective evaluation function, which comprehensively considers at least the stability improvement potential of the solution, the increase in execution cost, and the correlation impact on other operating parameters;

[0076] S4.4 Calculate the evaluation function score of each candidate scheme, sort them from high to low according to the evaluation function score, output the scheme with the highest score as the optimal coal-fired mixing scheme, and use the subsequent schemes as the backup scheme sequence.

[0077] In this embodiment, each step further includes:

[0078] Analyze the operating parameter deviations generated in step S3, and identify the parameter that poses the greatest threat to the safe and stable operation of the unit or has the most serious deviation as the most important operating deviation item.

[0079] Based on the physical meaning of the most important operational deviation, determine its correlation with the quality of the coal. For example, if the main steam temperature is consistently low, prioritize screening those mixed schemes that use coals with higher volatile matter or higher calorific value. Conversely, if the furnace fluctuates drastically, prioritize excluding mixed schemes with excessively high ash content or significantly different combustion characteristics.

[0080] For candidate solutions that pass the initial screening, a multi-factor evaluation system is established to rank them. This system includes at least three dimensions: the first dimension is the expected effect of the solution on correcting the most significant operational deviations, with a higher score for better effect; the second dimension is the economic efficiency of the solution, i.e., the difference between the implementation cost of the solution and the current cost, with a higher score for smaller cost increment; the third dimension is the robustness of the solution, assessing the potential risk of the solution adversely affecting other operational parameters after implementation, with a higher score for lower risk.

[0081] Each dimension's evaluation result is assigned a weight coefficient, and the comprehensive score of each candidate scheme is calculated by weighting. The schemes are sorted from high to low according to their comprehensive scores, and the scheme ranked first is output as the optimal coal-fired mixed scheme, while the schemes ranked second and third are used as backup schemes.

[0082] Furthermore, referring to Figure 4 Step S6, which involves establishing the dynamic safe operating range of the key equipment status parameters under the deep peak shaving condition, specifically includes:

[0083] S6.1 For each type of critical equipment, a standard safe operating range for the critical equipment under rated load is preset;

[0084] S6.2 Establish a load-safety interval correction coefficient mapping table and define the correction coefficients for the upper and lower limits of the standard safe operating interval of each device under different load levels;

[0085] S6.3 Establish a coal quality characteristic-safety range influence factor mapping table to define the direction and magnitude of the influence of different key coal quality characteristic indicators on the standard safe operating range of specific equipment;

[0086] S6.4. Based on the current actual load of the unit, query the load-safety interval correction coefficient mapping table and make the first correction to the standard safe operating interval;

[0087] S6.5. Based on the key coal quality characteristic indicators of the currently implemented or soon-to-be-implemented coal-fired mixed combustion scheme, query the coal quality characteristic-safety range influence factor mapping table, and perform a second correction on the safe operating range after the first correction to obtain the dynamic safe operating range under the current operating conditions.

[0088] In this embodiment, each step further includes:

[0089] During system initialization, the upper and lower limits of static safe operation under the rated working conditions specified by the manufacturer are entered for the status parameters of each type of critical equipment.

[0090] Construct a "load-safety factor" lookup table. This table determines the extent to which equipment status parameters are allowed to deviate from their rated static limits under different load rates (e.g., from deep load to rated load). The table is a two-dimensional table, with the horizontal axis representing the unit load percentage and the vertical axis representing different equipment. The table content is the safety boundary scaling factor corresponding to that load (e.g., the lower the load, the more lenient the upper limit of the allowable turbine current may be).

[0091] A "Coal Quality-Influence Coefficient" lookup table is constructed. This table is established by analyzing the impact of different coal quality characteristics on the operating load of specific equipment. For example, when the received ash content of mixed coal increases, the grinding load of the coal mill increases, and the safe operating limit of its motor current should be adjusted downward by an offset accordingly. This table is also a two-dimensional table. The horizontal axis represents the numerical range of key coal quality characteristic indicators (such as ash content and moisture), the vertical axis represents different equipment, and the table content is the offset or offset coefficient of the safety boundary.

[0092] Perform the first correction (load correction), read the current actual load value of the unit, query the "load-safety factor" lookup table, find the scaling factor of the equipment status parameter under the corresponding load; multiply the rated static safety upper limit by the scaling factor to obtain the upper limit after load correction, and multiply the rated static safety lower limit by the scaling factor to obtain the lower limit after load correction.

[0093] Perform the second correction (coal quality correction), read the dominant coal quality characteristic value of the current coal-fired mixing scheme, query the "coal quality-influence coefficient" lookup table, find the offset of the safety boundary of the equipment state parameter under the corresponding coal quality value; add (or subtract) this offset to the upper limit after load correction to obtain the upper limit of the final safe operating range after coal quality correction; similarly, offset the lower limit to obtain the lower limit of the final safe operating range.

[0094] Furthermore, in step S7, the first preset duration is set differently according to different key equipment status parameters.

[0095] In this embodiment, in step S7, the length of the first time window is set according to the parameter characteristics: for parameters with a fast rate of change, the first time window is set to be shorter; for parameters with a large change inertia, the first time window is set to be longer, in order to avoid false alarms.

[0096] Furthermore, in step S7, when a Level 1 warning signal is generated, a list of one or more suspected causes that lead to the abnormality of the parameter is output synchronously.

[0097] In this embodiment, in step S7, when a first-level warning signal is generated, the system simultaneously outputs a list of inferred causes. This list is obtained by querying the correlation between similar parameter anomalies and operational changes and coal quality changes in the historical case database.

[0098] Furthermore, the interlocking recommendation or execution of corresponding equipment maintenance or system operation instructions in step S8 includes:

[0099] It is recommended to switch to a coal-fired mixed scheme in the backup plan sequence, reduce the output of the corresponding coal mill, or increase the frequency of inspections of specific equipment.

[0100] In this embodiment, step S8 interlocks and triggers a predefined set of operation suggestions, specifically including: if the secondary alarm is caused by an excessively high current in a coal mill, the operation suggestion is "Execute immediately: reduce the set value of the coal feeder speed of the coal mill by 10%"; if the secondary alarm is caused by an excessively high bearing temperature in a primary air fan, the operation suggestion is "It is recommended to switch to the backup coal combustion scheme B first and strengthen the inspection of the cooling system of the fan".

[0101] Furthermore, the key coal quality characteristics mentioned in step S6.3 include, but are not limited to, volatile matter, ash content, moisture content, and ash fusion point on an as-received basis.

[0102] In this embodiment, the key coal quality characteristics in step S6.3 include volatile matter, ash, moisture and grindability coefficient on an as-received basis.

[0103] Furthermore, the pulverizing system parameters mentioned in step S2 include the coal mill outlet temperature, coal mill differential pressure, and primary air volume.

[0104] In this embodiment, the parameters of the pulverizing system in step S2 also include the current of the coal mill motor.

[0105] Furthermore, the method also includes step S9: recording each scheme adjustment instruction, early warning signal, alarm signal, and subsequent operation and system response, and forming a case library for periodic optimization and updating of the typical coal quality characteristic database, scheme generation rules, and dynamic safe operation range calculation rules.

[0106] In this embodiment, data from the entire process of scheme generation and adjustment to early warning and alarm is recorded, including the boundary conditions that trigger the decision, the final executed scheme, the content of the early warning and alarm, and subsequent manual intervention measures. This record forms a historical case library, which is used by engineers to periodically review and calibrate the parameters of the typical coal quality characteristic database, the weight coefficients in the scheme generation logic, and the lookup table data in the dynamic safe operation range calculation.

[0107] The following is a specific example:

[0108] This embodiment demonstrates in detail the complete application process of this method when a large coal-fired power generating unit participates in deep peak shaving. The entire process is implemented through the system's built-in logical rules, database, and predefined lookup tables, without involving any specific numerical values, algorithm code, or mathematical models.

[0109] Background: The power plant uses various types of coal with significant differences in quality. The units need to frequently respond to grid dispatch instructions and operate within a wide load range, especially under deep peak-shaving conditions at low loads, posing significant challenges to boiler combustion stability, steam parameter control, and auxiliary equipment safety.

[0110] Initial scheme generation:

[0111] Triggering and Data Acquisition: One afternoon, the power grid load demand decreased, and the dispatching instruction required the generating units to reduce their output to a target load far below their rated value within a specified time. The decision-making system was immediately activated. First, it acquired and parsed the target load instruction through the power plant's Information System for Monitoring and Control (SIS). Subsequently, the system queried the fuel management subsystem to obtain real-time coal level data for each raw coal bunker, accurately grasping the current inventory of available coal types (assuming they are coal A, coal B, and coal C). Simultaneously, the system retrieved detailed characteristic records of these coal types from its integrated "Typical Coal Quality Characteristics Database," including but not limited to key parameters such as industrial analysis (moisture, ash, volatile matter), elemental analysis, calorific value, ash fusion point, and Hardgrove grindability index for each coal type.

[0112] Scheme Calculation and Generation: The system takes meeting the total heat demand under the target load as its fundamental constraint. Based on this, it initiates the blending calculation logic with the core optimization direction of minimizing the expected total coal procurement cost. The system automatically enumerates various feasible coal combinations and blending ratios (e.g., A-type coal as the main component and B-type coal as a supplement, balanced blend of B-type and C-type coal, C-type coal as the main component and A-type coal as a supplement, etc.). For each hypothetical blending scheme, the system queries the database and uses a weighted average method to calculate the theoretical received basis lower heating value, received basis volatile matter, and received basis sulfur content of the blended coal, among other core indicators.

[0113] Safety Screening and Expected Setting: The system incorporates a range of coal quality standards allowed by the unit design. All calculated theoretical values ​​for the mixed coal quality are compared to this standard range. Any schemes that could lead to excessive sulfur content or excessively low volatile matter in the mixed coal, potentially jeopardizing combustion stability, are automatically screened and eliminated. Ultimately, the system successfully generates several (e.g., three) technically feasible and economically superior initial coal mixing schemes. For each scheme, expected control ranges for key operating parameters such as boiler main steam pressure, main steam temperature, and reheat steam temperature are preset under target load conditions.

[0114] Operating parameter monitoring:

[0115] The unit began to reduce load as instructed. During this period, the decision-making system maintained high-speed, uninterrupted communication with the unit's distributed control system (DCS) through a data interface, collecting and monitoring a large number of operating parameters in real time.

[0116] Boiler-side parameters include main steam pressure, main steam temperature, reheat steam temperature, furnace negative pressure, and flue gas oxygen content. These parameters directly reflect the boiler's combustion efficiency, heat exchange status, and operational stability.

[0117] Pulverizing system parameters include the outlet temperature of each operating coal mill, the differential pressure between the inlet and outlet of the coal mill, the primary air volume, and the motor current of the coal mill. These parameters are used to monitor the operating status and output of the pulverizing system, with the motor current being a key indicator that directly reflects the grinding load of the coal mill.

[0118] Plan adjustment judgment:

[0119] After the unit load drops to the target value and operates stably for a period of time, the decision-making system continuously compares the real-time operating parameters monitored in step 2 with the expected parameter range corresponding to the currently executing initial plan (assuming plan 1: coal A as the main component and coal B as the auxiliary component).

[0120] The system detected a sustained, slow decrease in the measured main steam temperature, gradually deviating from the lower limit of the expected range preset in Scheme 1. The system's built-in dead-zone threshold function was activated to ignore minor normal fluctuations. However, this deviation exceeded the dead-zone threshold. The system started an internal timer to monitor the duration of this exceeding condition. When the duration reached a preset judgment period (a time length set according to process characteristics), the system logic determined that the current operating condition deviated significantly from the expected value, formally triggering the "hybrid scheme dynamic adjustment command."

[0121] Dynamic optimization of the solution:

[0122] Deviation identification: After receiving the adjustment command, the system first analyzes all parameters that have deviated and determines that "the main steam temperature is consistently too low" is the core deviation that has the greatest impact on the current unit's economy and safety and needs to be addressed first.

[0123] Preliminary screening: Based on thermodynamic principles and an operational experience knowledge base, the system determines that increasing the main steam temperature requires enhanced radiative heat absorption within the furnace. Therefore, it re-evaluates all initial schemes generated in step 1. Scheme 1 (the current scheme) has proven ineffective. Another scheme (hypothetically, Scheme 2) is initially excluded because it incorporates high-ash coal, which the system logic determines may exacerbate the risk of slagging within the furnace, potentially further deteriorating heat exchange.

[0124] Comprehensive Assessment and Ranking: For the remaining candidate schemes (mainly Scheme 3: primarily C coal with A coal as a supplement), the system initiates a multi-factor comprehensive assessment process. Assessment dimensions include: a) Corrective Effect: Scheme 3 uses high-calorific-value coal as the main component, resulting in a high theoretical combustion temperature and an expected effective increase in steam temperature; this aspect receives a high score. b) Economic Efficiency: Scheme 3 has a higher cost; this aspect receives a lower score. c) Robustness: The system assesses the potential impact of Scheme 3 on other systems (such as the pulverizing system and environmental protection facilities), and determines the risk is controllable; this aspect receives a moderate score.

[0125] Decision Output: The system assigns pre-defined weight coefficients to each evaluation dimension (correction effect has the highest weight). After weighted calculation, Scheme 3 has the highest overall score and is selected as the "optimal coal-fired hybrid scheme." Scheme 1 is listed as the primary backup scheme due to its cost advantage. The system outputs the decision results to the operators' interface and provides implementation suggestions.

[0126] Equipment status monitoring:

[0127] The operators confirmed and implemented the system-recommended Plan 3, increasing the feed rate of C coal. Simultaneously, the decision-making system intensified monitoring of the specific coal mills and their auxiliary equipment related to grinding C coal.

[0128] The key parameters monitored include: the motor bearing temperature of the coal mill (reflecting mechanical load and lubrication condition), motor current (reflecting electrical load and grinding output), vibration value (reflecting mechanical stability), as well as the motor current and bearing temperature of the corresponding primary air fan (reflecting the air supply system load), and the speed feedback and casing temperature of the coal feeder (reflecting the smoothness of coal feeding).

[0129] Dynamic range setting:

[0130] The system has detected that the current operating conditions have changed: the load is in a deep peak shaving state and the dominant coal type has been switched to C coal (which is characterized by high calorific value, low volatile matter, and low Hardgrove grindability, i.e., difficult to grind).

[0131] The system sets a dynamic safe operating range for the current of the coal mill motor, which is currently under close monitoring. This process consists of two steps:

[0132] Load Correction: The system queries the preset "Load-Safety Factor Mapping Table". This table, based on historical operating data and equipment characteristics, indicates that under the current ultra-low load conditions, the allowable upper limit of the coal mill motor current can be relaxed compared to the static safety value under rated conditions (because the overall output requirement decreases). The system performs the first correction based on the corresponding coefficients in the table.

[0133] Coal Quality Correction: Next, the system queries the preset "Coal Quality-Influence Factor Mapping Table." This table defines the impact of different coal quality characteristics on equipment load. For the low Hardgrove grindability coefficient of coal C, the table indicates that a negative offset correction is needed for the upper limit of the safe current for this coal mill motor (because grinding is more difficult, the motor load is greater, and the safe limit needs to be reduced to prevent overload). The system performs a second correction based on the offset in the table.

[0134] After these two steps of correction, a tightened dynamic safe operating range was finally generated, specifically for the current "low load + difficult-to-grind coal" working condition.

[0135] Level 1 Warning:

[0136] The system compares in real time the coal mill motor current value monitored in step 5 with the dynamic safe operating range calculated in step 6.

[0137] It observed that the current value continuously increased due to the difficulty in grinding the coal, eventually reaching the upper limit of the dynamic range. The system uses a startup delay for judgment. For a parameter like current that changes rapidly, the system sets a short initial time window (e.g., tens of seconds). The current value remains near the upper limit of the range within this window duration.

[0138] The system immediately generated a Level 1 warning signal, which clearly stated: "Warning: The motor current of coal mill No. XX is continuously too high and is approaching the dynamic safe operating limit under the current conditions."

[0139] At the same time, the system calls up the historical case library to conduct causal analysis and adds the following prompt to the warning message: "Possible related factors: The current coal type has a low Hardgrove grindability coefficient; it is recommended to check: the coal mill loading force setting and the wear condition of the grinding parts."

[0140] Level 2 alarm and interlocking actions:

[0141] After receiving the warning, the operators began to check the situation. However, because C-type coal was indeed difficult to grind, the motor current continued to rise slowly but steadily, eventually exceeding the dynamic safety range and reaching the system's preset, more stringent emergency action threshold.

[0142] The system immediately upgraded the alarm level to a Level 2 fault alarm, with a more urgent alarm message: "Critical: The motor current of coal mill No. XX has exceeded the limit, posing a risk of overload and burnout!"

[0143] At the same time, the system automatically interlocks and triggers a set of pre-defined operation suggestions for such faults. The suggestions are very specific and actionable, such as: "Immediate action: Reduce the setpoint of the feeder speed corresponding to this coal mill by 10 percent" and "Long-term suggestion: Consider switching to backup plan 1 (mainly A coal) and plan to inspect and replace the grinding components of this coal mill".

[0144] Case study library learning and optimization:

[0145] Once all deep peak shaving tasks are completed and events are processed, the system will automatically initiate the record archiving process.

[0146] It encrypts and stores all the data from this event, including the initial target load, generated alternative solutions, decision-making basis, solution execution process, parameter deviation details, early warning and alarm triggering conditions and times, operational measures taken by operators, and the final system response results, in the historical case database.

[0147] These detailed case records will be regularly reviewed and analyzed by the power plant's professional engineering team for the following purposes: a) verifying and updating the accuracy of parameters for relevant coal types in the "Typical Coal Quality Characteristics Database"; b) optimizing the weight allocation and screening rules in the scheme generation logic; and c) calibrating the correction coefficients and offsets in the "Load-Safety Factor Mapping Table" and the "Coal Quality-Influence Factor Mapping Table." This ensures that the judgments of the entire decision-making system become increasingly accurate and reliable with the accumulation of practical experience.

[0148] This extended embodiment describes in great detail every logical step and decision-making stage of the method, from instruction response to final learning and optimization. The entire process demonstrates how intelligent collaborative decision-making and proactive early warning of equipment failures in coal-fired power plants can be achieved through non-algorithmic, rule-based, and knowledge-based methods.

[0149] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0150] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

[0151] The specific embodiments of the invention have been described in detail above, but they are only examples, and this application is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this application. Therefore, all equivalent changes, modifications, and improvements made without departing from the spirit and principles of this application should be covered within the scope of this application.

Claims

1. A method for deep peak shaving of coal-fired power units using AI-assisted collaborative decision-making and fault early warning of coal-fired power plant hybrid systems, characterized in that, Includes the following steps: S1. Based on the target load command and the current status of the unit, and using a pre-set database of typical coal quality characteristics, generate multiple initial coal mixing schemes. Specifically, based on the target load command, determine the target heat load requirement of the unit; based on the target heat load requirement and the available coal types and stock of the current unit, select at least two available basic coal types from the database of typical coal quality characteristics; with the optimization objective of meeting the target heat load requirement and minimizing total cost, calculate the theoretical values ​​of the mixed coal quality of each basic coal type under different proportions, and generate the multiple initial coal mixing schemes and their corresponding expected parameter ranges. S2. Real-time monitoring of the unit's operating parameters under deep peak shaving conditions, including boiler main steam pressure, main steam temperature, reheat steam temperature, furnace negative pressure, flue gas oxygen content, and pulverizing system parameters. S3. Compare the operating parameters with the expected parameter range of the currently executed coal-fired power generation scheme. If the operating parameters continue to deviate from the expected parameter range, trigger a dynamic adjustment command for the power generation scheme. S4. In response to the dynamic adjustment command of the mixing scheme, the multiple initial coal-fired mixing schemes are screened and sorted based on real-time operating parameters, and the optimal coal-fired mixing scheme and the sequence of backup schemes are output. Specifically, the real-time operating parameters monitored in step S2 are obtained, and the most significant operational deviation is identified. Based on the most significant operational deviation, schemes expected to exacerbate the deviation are removed from the multiple initial coal-fired mixing schemes. For the remaining candidate schemes, a multi-objective evaluation function is constructed, which comprehensively considers at least the stability improvement potential of the scheme, the increase in execution cost, and the correlation impact on other operating parameters. The evaluation function score of each candidate scheme is calculated, and the schemes are sorted from high to low according to their evaluation function scores. The scheme with the highest score is output as the optimal coal-fired mixing scheme, and the subsequent schemes are used as the sequence of backup schemes. S5. Execute the optimal coal mixing scheme and continuously monitor the status parameters of key equipment, including the temperature of the coal mill bearing, the current of the primary air fan, and the feedback of the coal feeder speed. S6. Establish the dynamic safe operating range of the key equipment status parameters under the deep peak shaving condition. The upper and lower limits of the dynamic safe operating range are dynamically adjusted according to the changes in the random group load and coal-fired mixing scheme. For each type of key equipment, a standard safe operating range of the key equipment under rated load is preset. A load-safe range correction coefficient mapping table is established to define the correction coefficients for the upper and lower limits of the standard safe operating range of each equipment under different load levels. A coal quality characteristic-safe range influence factor mapping table is established to define the direction and magnitude of the influence of different key coal quality characteristic indicators on the standard safe operating range of specific equipment. Based on the current actual load of the unit, the load-safe range correction coefficient mapping table is queried to perform the first correction on the standard safe operating range. Based on the key coal quality characteristic indicators of the currently executed or soon-to-be-executed coal-fired mixing scheme, the coal quality characteristic-safe range influence factor mapping table is queried to perform the second correction on the safe operating range after the first correction, thus obtaining the dynamic safe operating range under the current operating condition. S7. Real-time comparison of the status parameters of the key equipment with their corresponding dynamic safe operating range. When any parameter exceeds its dynamic safe operating range and continues for a first preset time, a first-level warning signal is generated. S8. After a Level 1 warning signal is generated, if the parameter further deteriorates to the point of exceeding the preset emergency threshold outside its dynamic safe operating range, a Level 2 fault alarm signal is generated, and the corresponding equipment maintenance or system operation instructions are recommended or executed.

2. The method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on deep coal-fired power plant peak shaving, as described in claim 1, is characterized in that... In step S7, the first preset duration is set differently according to different key equipment status parameters.

3. A method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on coal-fired power plant hybrid systems, as described in claim 1 or 2, characterized in that... In step S7, when a Level 1 warning signal is generated, a list of one or more suspected causes that could lead to the abnormality of the parameter is output synchronously.

4. The method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on coal-fired power plant hybrid systems according to claim 1, characterized in that, The interlocking recommendation or execution of the corresponding equipment maintenance or system operation instructions in step S8 includes: It is recommended to switch to a coal-fired mixed scheme in the backup plan sequence, reduce the output of the corresponding coal mill, or increase the frequency of inspections of specific equipment.

5. The method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on coal-fired power plant hybrid systems according to claim 1, characterized in that, The key coal quality characteristics mentioned in step S6.3 include, but are not limited to, volatile matter, ash content, moisture content, and ash fusion point on an as-received basis.

6. The method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on deep coal-fired power plant peak shaving, as described in claim 1, is characterized in that... The pulverizing system parameters mentioned in step S2 include the pulverizer outlet temperature, pulverizer differential pressure, and primary air volume.

7. The method for deep peak shaving of coal-fired power units using AI collaborative decision-making and fault early warning based on coal-fired power plant hybrid systems according to claim 1, characterized in that, The method further includes step S9: recording each scheme adjustment instruction, early warning signal, alarm signal, and subsequent operation and system response, and forming a case library for periodic optimization and updating of the typical coal quality characteristic database, scheme generation rules, and dynamic safe operation range calculation rules.

Images