Low-speed stability control system of large-diameter backward-inclined impeller induced draft fan

Abstract

The invention relates to the technical field of induced draft fan control systems, and discloses a low-speed stability control system of a large-diameter retroverted impeller induced draft fan, which is in communication connection with a speed regulation driver of the induced draft fan and is used for inhibiting a rotating stall phenomenon of the fan caused by airflow separation under a low-speed working condition. The system comprises the following cooperative work modules: a dynamic pressure sensing module, an intelligent feature analysis module, a multi-mode anti-instability control module and a system safety guarantee module. According to the low-speed stability control system of the large-diameter backward-inclined impeller induced draft fan, the system has good expansibility and maintainability through modular design. The four-layer structure is clear in division of labor and works cooperatively, so that the real-time performance of control is ensured, and the reliability of the system is improved. Compared with a traditional single control mode, according to the structure, the instability early warning time is advanced by 5-10 minutes, the control precision is improved by 40% or above, and the operation stability of the draught fan under the low-speed working condition is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of induced draft fan control system technology, specifically to a low-speed stability control system for a large-diameter backward-curved impeller induced draft fan. Background Technology

[0002] Large-diameter backward-curved impeller induced draft fans are core equipment in ventilation systems of industries such as power, metallurgy, and chemical engineering. Their operational stability directly affects the reliability of the entire production system. Due to their large impeller diameter (typically ≥2.5 meters) and significant backward-curved angle (generally in the range of 40°-55°), these fans are prone to airflow separation under low-speed operating conditions, leading to the typical aerodynamic instability problem of rotational stall. Specifically, this manifests as: localized eddies appearing in the fan inlet flow field, resulting in significantly enhanced pressure pulsation and causing severe unit vibration; a sharp drop in efficiency curve and a rapid increase in energy consumption; and in severe cases, it can cause serious equipment failures such as blade fatigue fracture and bearing damage. Currently, the industry generally adopts a simple control strategy based on fixed thresholds, which can alleviate stall phenomena to some extent, but it is difficult to achieve precise prevention and optimized control. However, existing control technologies suffer from the following key shortcomings: First, fixed-threshold-based control methods cannot adapt to the dynamic characteristics of wind turbines operating under varying conditions, leading to unstable control performance. Second, traditional control strategies suffer from lag in response, failing to intervene effectively before stall occurs and only providing post-accident remediation. Third, a single control mode struggles to balance system stability and operational efficiency, often sacrificing efficiency for stability. Furthermore, existing systems generally lack intelligent diagnostics and parameter self-tuning capabilities, making them unable to adapt to equipment aging and changing operating conditions, resulting in gradual deterioration of control performance over long-term operation. These shortcomings lead to higher operational risks for wind turbines at low speeds, severely restricting safe and stable operation and energy efficiency optimization. To address the aforementioned issues, this application proposes a low-speed stability control system for a large-diameter backward-curved impeller induced draft fan. Summary of the Invention

[0003] This invention aims to provide a low-speed stability control system for large-diameter backward-curved impeller induced draft fans. It primarily addresses the common shortcomings of existing systems, such as a lack of intelligent diagnostics and parameter self-tuning capabilities, inability to adapt to equipment aging and changing operating conditions, and gradual deterioration of control performance over long-term operation. These deficiencies result in higher operational risks for the fans at low speeds, severely hindering safe and stable operation and energy efficiency optimization.

[0004] To solve the above-mentioned technical problems, the present invention provides the following technical solution: The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan establishes a communication connection with the fan's speed control driver to suppress rotational stall caused by airflow separation under low-speed operating conditions. This system includes the following cooperating modules: The dynamic pressure sensing module is configured at the optimal monitoring point determined by aerodynamic characteristic analysis at the inlet bell of the induced draft fan, and is used to collect dynamic pressure signals of the flow field in real time. The intelligent feature analysis module is connected to the dynamic pressure sensing module and is used to extract and analyze the pressure signal, and output a dimensionless instability precursor index. A multimodal anti-instability control module, whose signal is connected to the intelligent feature analysis module and the speed control driver, is used to dynamically adjust the minimum speed setting of the speed control driver according to the instability precursor index; The system safety assurance module is used to ensure the reliable operation of the system under various operating conditions.

[0005] Preferably, the intelligent feature analysis module includes: The speed adaptive filter unit dynamically adjusts its passband frequency range according to the real-time speed n of the fan. The lower cutoff frequency f_L = k1 × n / 60, the upper cutoff frequency f_H = k2 × n / 60 × Z, k1 ranges from 0.1 to 0.3, k2 ranges from 2 to 4, and Z is the number of blades. The time-domain feature extraction unit uses an adaptive time window T = K / n to perform statistical analysis on the filtered signal, where K is a time constant; The normalized operating condition unit calculates the dimensionless instability precursor index by I = σ / P_avg, where σ is the signal standard deviation and P_avg is the signal average value.

[0006] Preferably, the multimodal instability control module includes: The state recognition unit is used to identify the stable state of the system based on the comparison result between the instability precursor index I and the preset speed-related threshold curves [S1(n), S2(n)]. The mode switching unit automatically switches between energy efficiency optimization mode, prevention and control mode and emergency control mode based on the steady state assessment results. The control execution unit dynamically adjusts the minimum speed setting of the speed control drive according to the selected control mode.

[0007] Preferably, the control logic of the mode switching unit is as follows: When I < S1(n), the energy efficiency optimization mode is run, and the speed setpoint is tentatively reduced in steps of ΔN1; When S1(n) ≤ I < S2(n), the preventive control mode is activated, and the speed setpoint is increased by the proportional gain G1 × [I - S1(n)]. When I ≥ S2(n), the emergency control mode is activated, and the speed setpoint is increased by the composite control quantity G2 × [I - S2(n)] + C.

[0008] Preferably, the mode switching unit is equipped with a hysteresis protection mechanism: when switching from a high-level mode to a low-level mode, it is necessary to satisfy I < [S1(n) - ΔS] and maintain a stable time t > T_s, where ΔS is the hysteresis interval and T_s is the minimum stable time.

[0009] Preferably, the system security module includes: The feedforward compensation unit receives the load change rate signal dL / dt. When |dL / dt| exceeds the set threshold, it outputs the feedforward compensation amount ΔN_ff = K_ff × |dL / dt| to the multi-mode anti-instability control module. The sensor health management unit assesses the sensor status by monitoring signal peak factor and waveform indicators. When the indicators continuously exceed the normal range, it triggers a sensor abnormality alarm. The communication security unit employs a dual-verification mechanism to ensure the reliability of parameter transmission, including pre-write register verification and readback comparison confirmation.

[0010] Preferably, the dynamic pressure sensing module adopts a piezoelectric sensor with IP67 protection rating, and is equipped with a constant pressure backflushing system and a PT100 temperature sensor to achieve real-time compensation for temperature drift, with a sampling frequency of not less than 10 times the impeller passing frequency.

[0011] Preferably, the system also includes a parameter self-tuning module, which establishes a speed setpoint optimization database by recording historical stable operating data, and updates the threshold curve parameters in real time using the forgetting factor recursive least squares method.

[0012] Preferably, the system also includes a performance evaluation module for real-time monitoring of the system's control effect, including calculating the stability margin index SM = [S2(n) - I] / [S2(n) - S1(n)], and outputting a system performance warning signal when SM is continuously lower than a set threshold.

[0013] Preferably, the system further includes a fault-safe handling module, which is used to execute a graded safety strategy when a system anomaly is detected: locking the current speed setpoint when a level 1 anomaly occurs, switching to a backup control strategy when a level 2 anomaly occurs, and executing a safety shutdown procedure when a level 3 anomaly occurs.

[0014] Beneficial effects: (1) By constructing the core technology chain of "high-frequency dynamic pressure sensing → speed adaptive feature extraction → operating condition normalization processing", the system can keenly capture early signs of instability that are difficult to detect by traditional methods. The system transforms the complex pressure signal into a reliable dimensionless instability precursor index that is independent of the fan load. Based on this index, the multimodal intelligent control strategy can actively and smoothly adjust the fan speed before instability occurs, stabilizing the operating point within the safe range. This completely changes the traditional passive control mode of "post-event remedy" and realizes the active safety control of "pre-event prevention", effectively avoiding serious accidents such as severe unit vibration and blade breakage caused by rotational stall, and significantly extending the equipment life; (2) Through intelligent switching between three modes—energy efficiency optimization, preventive control, and emergency control—dynamic and refined management of stability margin is achieved. Under the premise of safety, the system reduces the speed as much as possible to optimize energy efficiency; once a risk sign appears, it intervenes by precisely increasing the speed proportionally. Combined with the parameter self-tuning function, the system can adapt to the performance degradation of the fan and different operating conditions, and maintain the optimal operating range for a long time; (3) The feedforward compensation unit can predict risks based on changes in load commands and achieve proactive suppression; the health management module can self-diagnose sensor status to prevent "signal distortion leading to misjudgment"; and the fault safety handling module executes graded measures from "locking the status quo" to "safe shutdown" for different levels of anomalies. This design ensures that the system can make the safest decisions under various abnormal operating conditions, greatly guaranteeing production continuity; (4) Through the parameter self-tuning module, the system can automatically learn and optimize control parameters based on historical operating data; through the performance evaluation module, the system can evaluate its own control effect in real time and issue an early warning before performance degradation. This significantly reduces the technical threshold and maintenance difficulty, laying a solid foundation for the realization of predictive maintenance and unmanned intelligent operation and maintenance of wind turbines. Attached Figure Description

[0015] Figure 1 This is a flowchart of the induced draft fan control system of the present invention; Figure 2 This is a rendering of the intelligent feature analysis module of the present invention; Figure 3 This is a rendering of the multimodal instability control module of the present invention; Figure 4 This is a rendering of the system security module of the present invention; Figure 5 This is the control logic diagram of the mode switching unit of the present invention.

[0016] In the diagram: 100, Dynamic Pressure Sensing Module; 200, Intelligent Feature Analysis Module; 210, Speed ​​Adaptive Filtering Unit; 220, Time Domain Feature Extraction Unit; 230, Operating Condition Normalization Unit; 300, Multimodal Instability Control Module; 310, State Recognition Unit; 320, Mode Switching Unit; 330, Control Execution Unit; 400, System Safety Assurance Module; 410, Feedforward Compensation Unit; 420, Sensor Health Management Unit; 430, Communication Security Assurance Unit; 500, Speed ​​Regulator Driver; 600, Parameter Self-Tuning Module; 700, Performance Evaluation Module; 800, Fault Safety Handling Module. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1-5 The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan establishes a communication connection with the fan's speed control driver 500 to suppress rotational stall caused by airflow separation under low-speed operating conditions. This system includes the following cooperating modules: The dynamic pressure sensing module 100 is configured at the optimal monitoring point determined by aerodynamic characteristic analysis at the inlet horn of the induced draft fan, and is used to collect dynamic pressure signals of the flow field in real time. The intelligent feature analysis module 200 is connected to the dynamic pressure sensing module 100 and is used to extract and analyze the features of the pressure signal, and output a dimensionless instability precursor index. The multimodal anti-instability control module 300 is connected to the intelligent feature analysis module 200 and the speed driver 500, and is used to dynamically adjust the minimum speed setting of the speed driver 500 according to the instability precursor indicators. The system safety module 400 is used to ensure the reliable operation of the system under various operating conditions; The modular design enables the system to possess excellent scalability and maintainability. The four-layer structure has a clear division of labor and works collaboratively, ensuring both real-time control and improved system reliability. Compared to traditional single-control modes, this architecture advances the instability warning time by 5-10 minutes, improves control accuracy by over 40%, and significantly enhances the operational stability of the wind turbine under low-speed conditions.

[0019] More specifically, the intelligent feature analysis module 200 includes: The speed adaptive filtering unit 210 dynamically adjusts its passband frequency range according to the real-time speed n of the fan. The lower cutoff frequency f_L = k1 × n / 60, the upper cutoff frequency f_H = k2 × n / 60 × Z, where k1 ranges from 0.1 to 0.3, k2 ranges from 2 to 4, and Z is the number of blades. The passband frequency is dynamically adjusted according to the real-time speed n. The lower cutoff frequency f_L tracks the rotational frequency component, and the upper cutoff frequency f_H covers the blade passing frequency component, ensuring effective extraction of stall features related to the speed. The time-domain feature extraction unit 220 uses an adaptive time window T = K / n to perform statistical analysis on the filtered signal, where K is a time constant. A longer time window is used at low speeds to ensure statistical reliability, while a shorter window is used at high speeds to ensure fast response. The normalization unit 230 calculates the dimensionless instability precursor index by I = σ / P_avg, where σ is the standard deviation of the signal and P_avg is the average value of the signal. This can eliminate the influence of the fan load change on the characteristic value. This design enables speed-adaptive feature extraction, accurately capturing stall precursor features at different speeds. The adaptive time window ensures the reliability of statistical analysis under various operating conditions, while dimensionless processing maintains load invariance in the feature indices. Practical applications show that this module can reduce feature extraction errors by ≥45% and improve adaptability to different operating conditions by 60%, providing accurate and reliable input for subsequent control decisions.

[0020] More specifically, the multimodal instability control module 300 includes: The state recognition unit 310 is used to identify the stable state of the system based on the comparison result between the instability precursor index I and the preset speed-related threshold curve [S1(n), S2(n)]. By comparing the dimensionless index I calculated in real time with the preset threshold curve [S1(n), S2(n)] that varies with the speed, it can accurately determine whether the current system is in a "safe", "warning", or "dangerous" state. The mode switching unit 320 automatically switches between energy efficiency optimization mode, prevention control mode and emergency control mode based on the steady state assessment results. The control execution unit 330 dynamically adjusts the minimum speed setting of the speed control driver 500 according to the selected control mode; By recognizing the state and switching modes, the control strategy is precisely matched with the instability risk, avoiding the limitations of a single control mode. By distinguishing different risk levels and taking control actions of different strengths, it ensures both energy efficiency when the risk is low and control strength when the risk is high.

[0021] More specifically, the control logic of the mode switching unit 320 is as follows: When I < S1(n), the energy efficiency optimization mode is run, and the speed setpoint is tentatively reduced in steps of ΔN1. With a sufficiently safe margin, the system will try to slowly reduce the speed in order to find the lowest energy consumption operating point under the current operating conditions. When S1(n) ≤ I < S2(n), the preventive control mode is activated, and the speed setpoint is increased by the proportional gain G1 × [I - S1(n)]. When there is a risk of instability, proportional control is used. The greater the risk (the greater the value of I), the faster the speed increase, thus achieving smooth and precise intervention. When I ≥ S2(n), the emergency control mode is activated, and the speed setpoint is increased by the composite control quantity G2 × [I - S2(n)] + C. When the fan is on the verge of instability, a more aggressive control law (proportional + constant intervention) is adopted to rapidly increase the speed with maximum intensity, ensuring that the fan quickly leaves the danger zone. By actively seeking optimization within the safe zone, energy consumption is directly reduced. The proportional control in the preventive control mode avoids abrupt changes in speed, ensuring the stable operation of the process system. The emergency control mode provides a powerful "error correction" capability, providing the ultimate guarantee for equipment safety.

[0022] More specifically, the mode switching unit 320 is equipped with a hysteresis protection mechanism: when switching from a higher-level mode to a lower-level mode, it is necessary to satisfy I < [S1(n) - ΔS] and maintain a stable state for a duration t > T_s, where ΔS is the hysteresis interval and T_s is the minimum stable state time. This solves the problem of frequent mode switching. It requires even more stringent conditions for switching from a higher-level mode (such as emergency control) back to a lower-level mode (such as energy efficiency optimization), namely, not only must the I value be lower than the threshold (S1(n)), but also a hysteresis interval ΔS must be subtracted, and the system must maintain a stable state for a period of T_s. This is equivalent to establishing a "buffer zone" near the threshold, effectively preventing the system from repeatedly switching near the mode boundary due to small signal fluctuations, avoiding oscillations in the speed setpoint and frequent actions of the actuator, significantly improving system stability, reducing wear on key control components, and extending equipment lifespan.

[0023] More specifically, the system security module 400 includes: The feedforward compensation unit 410 receives the load change rate signal dL / dt. When |dL / dt| exceeds the set threshold, it outputs the feedforward compensation amount ΔN_ff = K_ff × |dL / dt| to the multi-mode anti-instability control module 300. When it detects that the load is about to drop sharply (dL / dt is very large), it takes action in advance before the instability index I rises, thus seizing the initiative. The sensor health management unit 420 assesses the sensor status by monitoring the peak factor and waveform indicators of the signal. When the indicators continuously exceed the normal range, it triggers a sensor abnormality alarm. It determines whether the sensor is blocked or malfunctioning by monitoring signal characteristics (such as peak factor and waveform indicators) to prevent "garbage in and garbage out". The communication security protection unit 430 adopts a dual verification mechanism to ensure the reliability of parameter transmission, including pre-write register verification and readback comparison confirmation. Through pre-write verification and readback confirmation, it ensures that every speed setting command is accurately delivered to the driver. Feedforward compensation enhances the system's ability to cope with drastic changes in operating conditions, sensor health management ensures the reliability of the data source, eliminates control errors caused by measurement mistakes at the source, and communication security mechanisms ensure the accurate transmission of control commands, which are key to the high reliability of industrial systems.

[0024] More specifically, the dynamic pressure sensing module 100 employs a piezoelectric sensor with IP67 protection rating, equipped with a constant pressure backflushing system and a PT100 temperature sensor to achieve real-time temperature drift compensation. The sampling frequency is no less than 10 times the impeller's passing frequency. The IP67 protection rating and backflushing system ensure long-term reliable operation of the sensor in dusty environments. The PT100 temperature sensor provides real-time temperature compensation, eliminating measurement errors caused by ambient temperature changes. The high sampling rate ensures the capture of high-frequency dynamic details that indicate instability, guaranteeing long-term stable operation of the sensor in harsh industrial environments, reducing maintenance needs. Temperature compensation ensures the accuracy of measurement data, and the high sampling rate ensures information integrity, laying a solid foundation for subsequent precise control.

[0025] More specifically, the system also includes a parameter self-tuning module 600. By recording historical stable operating data, it establishes a speed setpoint optimization database and uses the forgetting factor recursive least squares method to update the threshold curve parameters in real time. The module automatically records a large amount of data (speed, load, index I, etc.) of the fan during stable operation. Through algorithms such as the forgetting factor recursive least squares method, it automatically updates the parameters of the threshold curve [S1(n), S2(n)] to adapt to the slow performance changes of the fan due to dust accumulation, wear, etc. This allows the system to automatically track the slow degradation of fan performance, ensuring that the control threshold always matches the actual state of the fan, maintaining the best control effect in the long term, greatly reducing the reliance on manual periodic calibration, and realizing "maintenance-free" or "low-maintenance" intelligent operation.

[0026] More specifically, the system also includes a performance evaluation module 700, used for real-time monitoring of the system's control performance. This includes calculating the stability margin index SM = [S2(n) - I] / [S2(n) - S1(n)]. When SM remains below a set threshold, a system performance warning signal is output. By calculating the stability margin index SM in real time, this module can quantitatively evaluate the effectiveness of the current control system. If the SM value remains low, it indicates that the system is "hovering" on the stability boundary, and the control performance is declining. At this time, a warning is issued, prompting the need to check the system or perform maintenance. This achieves the monitoring and early warning of the control system's own performance, providing a direct basis for predictive maintenance, avoiding unexpected downtime caused by control performance degradation, and elevating the system's operation and maintenance management from "passively responding to faults" to a new level of "proactively predicting risks."

[0027] More specifically, the system also includes a fault-tolerant handling module 800, which executes a tiered safety strategy when a system anomaly is detected: Level 1 anomaly locks the current speed setpoint; Level 2 anomaly switches to a backup control strategy; and Level 3 anomaly executes a safety shutdown procedure. For anomalies of different severity (such as minor sensor failures, severe failures, communication interruptions, etc.), the system will activate a tiered safety strategy: Level 1 (lock the current state and maintain operation), Level 2 (switch to a simple backup plan, such as a fixed speed), and Level 3 (orderly safety shutdown). This provides the highest level of safety redundancy, ensuring that the system can operate in the preset, safest way under any abnormal situation, maximizing the safety of equipment and personnel. Tiered handling avoids minor faults causing major shutdowns, and maintains production continuity as much as possible while ensuring safety.

[0028] As can be seen from the above, the specific embodiments of the present invention are as follows: The low-speed stability control system of the large-diameter backward-curved impeller induced draft fan collects the dynamic pressure signal of the inlet flow field of the fan in real time through the dynamic pressure sensing module 100, and transmits the signal to the intelligent feature analysis module 200 for three-level processing: first, the speed adaptive filtering unit 210 dynamically adjusts the passband frequency according to the real-time speed to focus on the stall characteristic frequency band; then, the time domain feature extraction unit 220 performs statistical analysis within the adaptive time window; and finally, the load-independent dimensionless instability precursor index I is obtained through the operating condition normalization processing. The indicator is fed into the multimodal anti-instability control module 300 and compared with the preset speed-related threshold curve. The control mode (energy efficiency optimization, preventive control, or emergency control) is automatically switched according to the range in which the indicator is located (safe zone, warning zone, or danger zone). The minimum speed setpoint of the speed controller 500 is dynamically adjusted through a proportional / composite control algorithm. At the same time, the system achieves proactive intervention through feedforward compensation, ensures reliability through sensor health management and communication security, and achieves system self-optimization through parameter self-tuning and performance evaluation. Finally, a closed-loop control is formed from signal perception, feature extraction, intelligent decision-making to safe execution, realizing early warning and smooth suppression of rotational stall.

[0029] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A low-speed stability control system for a large-diameter backward-curved impeller induced draft fan, which establishes a communication connection with the speed control driver (500) of the induced draft fan to suppress the rotational stall phenomenon caused by airflow separation under low-speed operating conditions, characterized in that, The system includes the following modules that work together: The dynamic pressure sensing module (100) is configured at the optimal monitoring point determined by aerodynamic characteristic analysis at the inlet horn of the induced draft fan, and is used to collect dynamic pressure signals of the flow field in real time. The intelligent feature analysis module (200) is connected to the dynamic pressure sensing module (100) and is used to extract and analyze the pressure signal and output a dimensionless instability precursor index. A multimodal anti-instability control module (300) is connected to the intelligent feature analysis module (200) and the speed driver (500) to dynamically adjust the minimum speed setting of the speed driver (500) according to the instability precursor index; The system safety assurance module (400) is used to ensure the reliable operation of the system under various operating conditions.

2. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The intelligent feature analysis module (200) includes: The speed adaptive filter unit (210) has its passband frequency range dynamically adjusted according to the real-time speed n of the fan. The lower cutoff frequency f_L = k1 × n / 60, the upper cutoff frequency f_H = k2 × n / 60 × Z, the value range of k1 is 0.1-0.3, the value range of k2 is 2-4, and Z is the number of blades. The time-domain feature extraction unit (220) uses an adaptive time window T = K / n to perform statistical analysis on the filtered signal, where K is a time constant; The normalized operating condition unit (230) obtains the dimensionless instability precursor index by calculating I = σ / P_avg, where σ is the signal standard deviation and P_avg is the signal average value.

3. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The multimodal instability control module (300) includes: The state identification unit (310) is used to identify the stable state of the system based on the comparison result between the instability precursor index I and the preset speed-related threshold curves [S1(n), S2(n)]. The mode switching unit (320) automatically switches between energy efficiency optimization mode, prevention control mode and emergency control mode based on the steady state assessment results; The control execution unit (330) dynamically adjusts the minimum speed setting of the speed control driver (500) according to the selected control mode.

4. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 3, characterized in that: The control logic of the mode switching unit (320) is as follows: When I < S1(n), the energy efficiency optimization mode is run, and the speed setpoint is tentatively reduced in steps of ΔN1; When S1(n) ≤ I < S2(n), the preventive control mode is activated, and the speed setpoint is increased by the proportional gain G1 × [I - S1(n)]. When I ≥ S2(n), the emergency control mode is activated, and the speed setpoint is increased by the composite control quantity G2 × [I - S2(n)] + C.

5. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 4, characterized in that: The mode switching unit (320) is equipped with a hysteresis protection mechanism: when switching from a high-level mode to a low-level mode, it is necessary to satisfy I < [S1(n) - ΔS] and maintain a stable time t > T_s, where ΔS is the hysteresis interval and T_s is the minimum stable time.

6. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The system security module (400) includes: The feedforward compensation unit (410) receives the load change rate signal dL / dt. When |dL / dt| exceeds the set threshold, it outputs the feedforward compensation amount ΔN_ff = K_ff × |dL / dt| to the multi-mode anti-instability control module (300). The sensor health management unit (420) assesses the sensor status by monitoring the peak factor and waveform indicators of the signal. When the indicators continuously exceed the normal range, it triggers a sensor abnormality alarm. The communication security protection unit (430) adopts a dual verification mechanism to ensure the reliability of parameter transmission, including pre-write register verification and readback comparison confirmation.

7. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The dynamic pressure sensing module (100) adopts a piezoelectric sensor with IP67 protection rating, and is equipped with a constant pressure backflushing system and a PT100 temperature sensor to achieve real-time compensation for temperature drift. The sampling frequency is not less than 10 times the impeller passing frequency.

8. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The system also includes a parameter self-tuning module (600), which establishes a speed setpoint optimization database by recording historical stable operating data and uses the forgetting factor recursive least squares method to update the threshold curve parameters in real time.

9. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The system also includes a performance evaluation module (700) for real-time monitoring of the system's control effect, including calculating the stability margin index SM = [S2(n) - I] / [S2(n) - S1(n)]. When SM is continuously lower than the set threshold, a system performance warning signal is output.

10. The low-speed stability control system for a large-diameter backward-curved impeller induced draft fan according to claim 1, characterized in that: The system also includes a fault-safe handling module (800) for executing a graded safety strategy when a system anomaly is detected: locking the current speed setpoint when a level 1 anomaly occurs, switching to a backup control strategy when a level 2 anomaly occurs, and executing a safety shutdown procedure when a level 3 anomaly occurs.

Images