A high tower granulation functional fertilizer production device

Abstract

The application discloses a high-tower granulation functional fertilizer production device, and relates to the technical field of fertilizer production. The device comprises a granulation tower body, an airflow adjusting assembly, a plurality of dynamic pressure detection units for obtaining dynamic pressure disturbance information generated by materials at different circumferential positions in the tower, and an acoustic detection assembly arranged at the bottom of the tower body and used for obtaining acoustic event information generated by the materials when the materials arrive at a bottom collecting area. A control assembly is configured to identify abnormal material transportation events in a middle area of the granulation tower body according to the dynamic pressure disturbance information, perform time correlation analysis on the abnormal material transportation events and the acoustic event information, and determine apparent transportation time of the materials from the middle area of the granulation tower body to the bottom collecting area. According to the apparent transportation time representing a material transportation state and circumferential pressure difference information between the plurality of dynamic pressure detection units, the airflow adjusting assembly is controlled to adjust airflow distribution in the tower body. The application has the effect of improving the quality of fertilizer output.

Description

Technical Field

[0001] This application relates to the technical field of fertilizer production, and in particular to a high-tower granulation functional fertilizer production device. Background Technology

[0002] Functional fertilizer is short for functional fertilizer, which is a widely used granular fertilizer. This fertilizer can be a fertilizer mixture formed by one or more fertilizer components, or it can be a functional fertilizer product formed by further adding organic matter, carrier or other additives to fertilizer components. It can play a role in improving soil and promoting plant growth.

[0003] High-tower granulation is a crucial process in the production of granular fertilizers. During granulation, the mixed granulation liquid is sprayed from the top of the tower and granulated into particles. These particles then settle to the bottom of the tower under gravity, simultaneously coming into counter-current contact with the cooling airflow flowing upwards from the bottom, achieving cooling and solidification. The actual settling velocity and residence time of the particles are not absolutely constant but are influenced by factors such as particle state, material load, and airflow distribution within the tower. Especially when the particles are still in a semi-fluid or incompletely solidified state with a certain viscosity, changes in local airflow distribution, material concentration, or cooling conditions within the tower can easily cause contact, agglomeration, or abnormal clustering between particles, thus affecting the uniformity and final quality of the functional fertilizer product.

[0004] Currently, monitoring methods for the forming and transporting states of materials within the high-tower granulation process are limited. Conventional methods often rely on sampling and analysis after the material exits the tower or on empirical judgment, which is a post-event detection method. This makes it difficult to detect and suppress abnormal agglomeration occurring within the tower in a timely manner, leading to batch-wide quality problems in fertilizer products. Conventional methods also employ online detection techniques such as optical detection, gamma rays, or observation windows. However, due to the complex production environment inside the tower, characterized by high temperature, high humidity, and high concentrations of fertilizer dust and gas-solid two-phase flow, these online detection methods have limitations under actual continuous production conditions. For example, optical detection is easily affected by fertilizer particles, dust, and moisture adhesion. Furthermore, high concentrations of gas-solid two-phase flow scatter and attenuate the light signal, significantly increasing the maintenance requirements for long-term online monitoring. While detection schemes using gamma rays or other radiation penetration methods can reduce the obstruction of conventional optical detection by high-concentration materials to some extent, they involve the safety management of radiation sources or devices, radiation protection, and corresponding on-site usage management requirements. Implementing these in continuous fertilizer production sites requires high levels of equipment management and operational expertise. As for observation windows or manual visual inspection, dust inside the tower tends to accumulate on the surface of the window, gradually reducing visibility. Furthermore, manual observation mainly relies on the operator's experience for qualitative judgment, making it difficult to continuously output quantitative transport status parameters that can be directly used for process control.

[0005] Therefore, existing alternatives are all unable to simultaneously ensure reliability, safety, and quantifiability under the special operating conditions of granulation towers. There is an urgent need for a material transport status monitoring solution that can achieve continuous, quantitative, and closed-loop monitoring without contacting the moving materials inside the tower. Summary of the Invention

[0006] To address the aforementioned technical problems, this application provides a high-tower granulation functional fertilizer production device.

[0007] This application provides a high-tower granulation functional fertilizer production device, which adopts the following technical solution:

[0008] A high-tower granulation functional fertilizer production device includes a granulation tower body, an air supply component for introducing cooling airflow into the granulation tower body, and an airflow regulating component for adjusting the airflow distribution inside the granulation tower body. The granulation tower body is used to granulate a fertilizer mixture composed of multiple fertilizer components and additives. The device also includes:

[0009] A dynamic pressure detection component is located in the middle of the granulation tower body and includes multiple dynamic pressure detection units distributed circumferentially along the granulation tower body. It is used to acquire dynamic pressure disturbance information generated by the gas-solid two-phase motion inside the granulation tower body at different circumferential positions.

[0010] An acoustic detection component is installed in the bottom collection area of ​​the granulation tower to acquire acoustic event information generated when the material reaches the bottom collection area;

[0011] The control component is electrically connected to the dynamic pressure detection component, the acoustic detection component, and the airflow regulation component, respectively. The control component is configured as follows:

[0012] Identify abnormal material transport events in the central region of the granulation tower based on dynamic pressure disturbance information;

[0013] By performing time correlation analysis between abnormal material transport events and acoustic event information, the apparent transport time of material moving from the middle area of ​​the granulation tower to the bottom collection area can be determined.

[0014] Based on the apparent transport time characterizing the material transport state and the circumferential pressure difference information between multiple dynamic pressure detection units, the airflow regulating component is controlled to adjust the airflow distribution inside the granulation tower.

[0015] Optionally, the control component can be configured as follows:

[0016] Acquire the dynamic pressure signals collected by each of the dynamic pressure detection units;

[0017] Establish dynamic pressure background values ​​corresponding to each detection location;

[0018] Based on the difference between the dynamic pressure signal and the dynamic pressure background value, a transient pressure disturbance signal that removes the influence of environmental pressure changes is obtained.

[0019] Abnormal pressure events are extracted based on the transient pressure disturbance signal.

[0020] Optionally, the control component can be further configured as follows:

[0021] The transient pressure disturbance signal is subjected to event feature extraction to obtain the occurrence time, disturbance amplitude, duration, and event energy parameters of the abnormal pressure event;

[0022] And based on the multiple abnormal pressure events, a sequence of abnormal transport events in the middle of the tower is formed.

[0023] Optionally, the control component is also configured as follows:

[0024] Acquire the bottom acoustic signal collected by the acoustic detection component;

[0025] Event recognition is performed on the bottom acoustic signal to obtain the material falling acoustic event sequence;

[0026] Cross-correlation analysis was performed on the abnormal transport event sequence in the middle of the tower and the acoustic event sequence of material falling, and the time delay corresponding to the peak value of the cross-correlation was taken as the apparent transport time.

[0027] Optionally, the control component is also configured as follows:

[0028] Acquire historical apparent transport time data under normal production conditions and establish a dynamic transport time benchmark.

[0029] The currently obtained apparent transport time is compared with the dynamic transport time benchmark to determine the degree of anomaly in the current material transport.

[0030] Optionally, the control component is also configured as follows:

[0031] Each of the multiple dynamic pressure detection units is acquired;

[0032] Based on the differences in pressure disturbance characteristic parameters corresponding to different circumferential positions, the distribution of abnormal material transport status in the circumferential direction of the granulation tower is determined, so as to control the airflow regulating component at the corresponding position to perform regulating action.

[0033] Optionally, the airflow regulation assembly includes multiple independently controlled airflow regulation units;

[0034] The control component is configured to determine the operating state of each airflow adjustment unit and execute corresponding adjustment actions based on the comparison result between the apparent transport time and the dynamic transport time reference, as well as the differences between the circumferential pressure disturbance characteristic parameters corresponding to the plurality of dynamic pressure detection units.

[0035] Optionally, the control component is also configured as follows:

[0036] After the airflow regulating component performs the regulating action, the dynamic pressure disturbance information and the acoustic event information are reacquired.

[0037] The apparent transport time and the circumferential pressure distribution state are re-determined based on the re-acquired dynamic pressure disturbance information and acoustic event information.

[0038] Based on the changes in material transport status before and after adjustment, the adjustment action of the airflow adjustment component is corrected.

[0039] In summary, this application includes at least one of the following beneficial effects:

[0040] 1. This application adopts a non-contact detection architecture that combines dynamic pressure detection located in the middle of the granulation tower body with acoustic detection located in the bottom collection area. The dynamic pressure detection component and the acoustic detection component work together to obtain the temporal correlation information of abnormal transport events in the middle and acoustic events of material falling at the bottom without contacting the material inside the tower. This provides quantifiable process input for closed-loop airflow regulation. Compared with conventional detection methods, this application does not require a continuously clean optical channel inside the tower, which can reduce the impact of fertilizer dust and water vapor on the detection results; it does not require the configuration of a radiation source or a corresponding radiation protection detection system, which reduces the special safety management requirements of the on-site detection system; and it does not rely on inefficient, experience-based and uncontinuous observation methods such as observation windows or manual visual inspection, which greatly reduces the adverse effects of high dust, high temperature and high humidity industrial environments on the detection elements, thereby achieving long-term reliable monitoring and automatic control of material transport status.

[0041] 2. This invention involves installing a circumferentially distributed dynamic pressure detection component in the middle of the tower body of a high-tower granulation device, and an acoustic detection component in the collection area at the bottom of the tower. The dynamic pressure detection component monitors pressure fluctuation signals at different heights and orientations within the tower in real time, constructing an abnormal pressure event sequence in the middle of the tower from the pressure fluctuation signals. The acoustic detection component captures acoustic characteristic signals generated by particle collisions, impacts, and friction, constructing an acoustic event sequence at the bottom of the tower from the acoustic characteristic signals. By performing cross-correlation analysis between the abnormal pressure event sequence in the middle and the acoustic event sequence at the bottom of the tower, the time delay corresponding to the correlation peaks of the two is used as the basis for the analysis. To achieve apparent transport time, the complex gas-solid two-phase motion state is transformed into a quantifiable transport time parameter. Through dynamic transport time benchmark and relative deviation calculation, the degree of material transport anomaly is quantitatively evaluated. At the same time, by combining the pressure disturbance characteristics differences of multiple circumferential detection points, the distribution of abnormal transport states in the circumferential direction of the tower body is identified, thereby accurately locating the areas where abnormal agglomeration or segregation occurs in the tower. This provides precise input for subsequent zone control. Abnormal transport information in the middle of the tower body can be obtained in a timely manner without direct contact with the material inside the tower. This overcomes the lag of traditional post-sampling and experience-based judgment and improves the timeliness of detecting abnormal material transport states.

[0042] 3. This invention controls multiple independent airflow adjustment units based on the degree of apparent transport time anomalies and circumferential pressure differences. It can adjust the intensity of local airflow in abnormal areas, avoiding the loss of uniformity caused by overall airflow adjustment. This more effectively suppresses particle agglomeration and abnormal aggregation, improves the solidification uniformity of functional fertilizer granules and product consistency. After airflow adjustment, it reacquires dynamic pressure disturbance information and acoustic event information, recalculates apparent transport time and circumferential pressure distribution, and corrects the airflow adjustment strategy based on the changes in state before and after adjustment. This enables the device to adapt to fluctuations in operating conditions and changes in abnormal states, continuously optimizes the transport state within the tower, avoids repeated anomalies, and greatly improves the granulation efficiency of fertilizer. Attached Figure Description

[0043] Figure 1 This is a cross-sectional schematic diagram illustrating the internal structure of the granulation tower body according to an embodiment of this application;

[0044] Figure 2 This is a schematic diagram illustrating the distribution of the dynamic pressure detection units in an embodiment of this application;

[0045] Figure 3 This is a cross-sectional schematic diagram illustrating the internal structure of the dynamic pressure detection unit in an embodiment of this application;

[0046] Figure 4 This is a cross-sectional schematic diagram illustrating the internal structure of the acoustic detection component in an embodiment of this application.

[0047] Explanation of reference numerals in the attached diagram: 1. Granulation tower body;

[0048] 2. Air supply components; 21. Zoned air supply ducts;

[0049] 3. Airflow regulating components; 31. Louvered air guides;

[0050] 4. Dynamic pressure detection unit; 41. Pressure tapping structure; 411. Measuring pressure tapping section; 412. Protective section; 413. Isolation section; 42. Reference pressure tapping assembly; 421. Reference pressure tapping port; 422. Damping connecting pipe; 423. Reference pressure chamber; 43. Differential pressure detection element;

[0051] 5. Acoustic detection component; 51. Acoustic pickup interface; 52. Acoustic waveguide cavity; 53. Flexible acoustic diaphragm; 54. Microphone. Detailed Implementation

[0052] 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 them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the protection scope of this application.

[0053] This application discloses a high-tower granulation functional fertilizer production device, see below. Figure 1 and Figure 2 The high-tower granulation functional fertilizer production device includes a granulation tower body 1, an air supply assembly 2, an airflow regulation assembly 3, a dynamic pressure detection assembly, an acoustic detection assembly 5, and a control assembly. The granulation tower body 1 forms a granulation space where material settles from top to bottom and cooling airflow flows from bottom to top. Both the air supply assembly 2 and the airflow regulation assembly 3 are located at the bottom of the granulation tower body 1. The air supply assembly 2 introduces cooling airflow from bottom to top into the granulation tower body 1. The airflow regulation assembly 3 regulates the overall flow rate and the circumferential flow rate of the cooling airflow entering the granulation tower body 1. The dynamic pressure detection assembly is located in the middle of the granulation tower body 1 to acquire local dynamic pressure disturbances caused by the movement of material in the middle area of ​​the tower body. The acoustic detection assembly 5 is located in the collection area at the bottom of the granulation tower body 1 to acquire acoustic events generated when material reaches the bottom collection area. The control component is electrically connected to the dynamic pressure detection component, the acoustic detection component 5 and the airflow adjustment component 3 respectively. It can determine the material transport status based on the detection information obtained from different height positions, and adjust the cooling airflow distribution inside the granulation tower 1 according to the material transport status and the circumferential pressure distribution inside the tower.

[0054] In this embodiment, the granulation tower 1 is used to granulate and cool the material of the functional fertilizer. This functional fertilizer is a mixture of fertilizer components, or it can be a mixture of fertilizer components and other added components such as organic matter and microbial components. The specific method can be determined according to the function of the functional fertilizer.

[0055] In some embodiments, the air supply assembly 2 includes a cooling fan, a main air supply duct, and multiple partitioned air supply ducts 21 connected to the main air supply duct. The cooling fan generates cooling airflow into the granulation tower 1, the main air supply duct collects and transports the cooling airflow, and the multiple partitioned air supply ducts 21 are respectively connected to different circumferential areas at the bottom of the granulation tower 1, so that the cooling airflow entering the granulation tower 1 can form independent airflow inputs in different circumferential areas. The airflow regulating assembly 3 is installed at the output end of each partitioned air supply duct 21. In this embodiment, the airflow regulating assembly 3 includes four sets of independently controlled electric louvered guides 31, which are evenly distributed along the circumference of the granulation tower 1. Each set of electric louvered guides 31 includes a rotatable louvered guide plate, a servo actuator connected to the louvered guide plate, and an angle position sensor for detecting the current position of the louvered guide plate. The control component can output target opening commands to the four sets of servo actuators respectively, so that the four sets of louvered guide plates can independently change the opening within a certain range, thereby changing the intensity and angle of the cooling airflow entering the granulation tower 1 in the corresponding circumferential area.

[0056] In this embodiment, the dynamic pressure detection component is located in the middle of the granulation tower 1. The specific height is determined based on the temperature change and solidification state of the functional fertilizer material within the granulation tower 1, and is preferably located in the lower part of the transport area where the granules are still in a certain viscous or semi-fluid state. The selection of the installation position of the dynamic pressure detection component does not require all granulation towers 1 to use a uniform fixed height. Instead, it is determined after prototype debugging based on the granulation temperature, solidification temperature, moisture content, and normal falling time of the specific material, enabling the dynamic pressure detection component to acquire dynamic disturbance information generated during the transport process before the material reaches the bottom of the tower.

[0057] In some embodiments, the dynamic pressure detection assembly includes three dynamic pressure detection units 4 positioned at the same horizontal height in the granulation tower 1, and evenly spaced along the circumference of the granulation tower 1. Preferably, the circumferential angle between the dynamic pressure detection units 4 is 120°. The three dynamic pressure detection units 4 divide the area at the same height in the granulation tower 1 into a first circumferential detection area, a second circumferential detection area, and a third circumferential detection area. The three detection positions do not operate independently but are used together to acquire dynamic pressure changes at different circumferential positions at the same height in the granulation tower 1, so that the control assembly can determine whether abnormal material transport is concentrated in a certain circumferential area and whether there is spatial non-uniformity in the gas-solid two-phase transport state within the tower.

[0058] Reference Figures 1 to 4 In this embodiment, each dynamic pressure detection unit 4 includes a pressure tapping structure 41, a reference pressure tapping component 42, a differential pressure detection element 43, and a signal acquisition module, all disposed on the side wall of the granulation tower 1. The pressure tapping structure 41 is installed at the detection position on the middle side wall of the granulation tower 1. Its side facing the inside of the tower forms an inwardly recessed measuring pressure tapping part 411 that communicates with the space inside the tower. The opening of the pressure tapping part does not directly face the direction of material falling, but a protective part 412 is provided on the outside of the opening to prevent material from entering the pressure tapping channel. The measuring pressure tapping part 411 is provided with a buffer chamber and an isolation part 413 with a baffle structure. The isolation part 413 with the baffle structure can transmit pressure disturbances inside the tower to the differential pressure detection element 43 through the pressure tapping part, and can also mitigate the situation where material dust enters the interior of the pressure tapping structure 41 and affects the accuracy of the detection element.

[0059] Furthermore, the reference pressure tapping component 42 is positioned on the tower wall of the granulation tower 1 at the same or similar height as the pressure tapping structure 41, avoiding the main material falling path. The reference pressure tapping structure 41 includes a reference pressure tapping port 421 disposed on the inner wall of the tower body, a damping connecting pipe 422 communicating with the reference pressure tapping port 421, and a reference pressure chamber 423 communicating with the damping connecting pipe 422. The reference pressure tapping port 421 does not directly face the main direction of material movement; its opening is set to be basically flush with the tower wall or adopts a recessed structure with a protective function, thereby reducing the possibility of particles directly impacting the reference pressure tapping port 421. The damping connecting pipe 422 adopts a relatively small flow cross-section and has a certain length, so that when the pressure change in the tower is transmitted to the reference pressure chamber 423 through the damping connecting pipe 422, flow damping and pressure buffering are generated, so that the pressure in the reference pressure chamber 423 mainly follows the slowly changing average pressure state in the tower, while having a strong attenuation effect on the high-frequency pressure disturbances generated by the rapid passage of local materials.

[0060] The differential pressure detection element 43 preferably employs a dynamic differential pressure sensor capable of responding to the target pressure change frequency range, such as a piezoresistive dynamic pressure sensor, a thin-film differential pressure sensor, or a MEMS differential pressure detection element 43. The specific model is selected based on the internal pressure change range and target frequency range of the granulation tower 1. The differential pressure detection element 43 is arranged inside the external insulation structure of the granulation tower 1 or in a protective box outside the insulation structure. The differential pressure detection element 43 has a measuring pressure interface and a reference pressure interface, wherein the measuring pressure interface is connected to the measuring pressure chamber, and the reference pressure interface is connected to the reference pressure chamber 423. Thus, the differential pressure detection element 43 can simultaneously acquire the local pressure tapping space pressure and the internal reference pressure of the tower, and output the dynamic pressure difference between the two.

[0061] Because changes in the overall air supply, ambient temperature, and slow pressure changes caused by the chimney effect within the granulation tower 1 can simultaneously affect both the measuring pressure side and the reference pressure side, these combined changes are weakened to some extent in the differential pressure detection results. Meanwhile, the rapid local pressure disturbance caused by abnormal material passing through the detection area corresponding to the pressure tapping structure 41 primarily affects the measuring pressure side. Due to the spatial location of the reference pressure tapping chamber and the damping effect of the damping connecting pipe 422, the reference pressure side does not experience a synchronous rapid change of the same amplitude. Therefore, it can form a relatively prominent transient pressure disturbance in the output signal of the differential pressure detection element 43. The signal acquisition module uses synchronous sampling to acquire the pressure signals output by the three dynamic pressure detection units 4 and provides a unified time reference for the three pressure channels.

[0062] In this embodiment, the sampling frequency of the dynamic pressure detection component is set according to the duration of the target transient pressure disturbance. In one embodiment, all three dynamic pressure detection units 4 use a 1kHz sampling frequency for synchronous sampling to simultaneously acquire pressure disturbance events with durations ranging from tens to hundreds of milliseconds. However, this sampling frequency is not intended to limit the implementation of this application. For specific devices, the sampling frequency can be increased or decreased according to the actual target frequency band, as long as the sampling frequency can meet the time-domain resolution requirements of the target dynamic pressure change.

[0063] Furthermore, the acoustic detection component 5 is installed in the bottom collection area of ​​the granulation tower 1. Because the dust concentration at the bottom of the tower is high, and the scraping and collecting mechanisms generate mechanical vibrations, in this embodiment, the acoustic detection component 5 does not directly insert the industrial microphone 54 into the granulation tower 1; instead, it employs an isolated acoustic pickup structure. Specifically, the acoustic detection component 5 includes an acoustic pickup interface 51 disposed on the side wall of the bottom collection area, an acoustic waveguide cavity 52 connected to the acoustic pickup interface 51, a flexible acoustic diaphragm 53 sealing the opening in the tower wall, and an industrial microphone 54 installed outside the granulation tower 1. The flexible acoustic diaphragm 53 isolates the air inside the tower from the space where the microphone 54 is located, preventing moisture and dust inside the tower from directly entering the microphone 54. Simultaneously, it allows the air pressure wave generated by the material impacting the bottom collection area to act on the flexible acoustic diaphragm 53, causing the flexible acoustic diaphragm 53 to undergo slight mechanical vibration and transmit acoustic information to the microphone 54. The microphone 54 is installed outside the acoustic waveguide cavity 52 through an elastic vibration isolator to reduce the impact of mechanical vibration of the granulation tower 1 and low-frequency structural noise generated by the operation of the scraping mechanism being directly transmitted to the microphone 54 through the solid path.

[0064] The control component can be an industrial computer, a PLC processing edge computing module, or an embedded controller with digital signal processing capabilities. This embodiment preferably uses an edge computing controller, enabling the three dynamic pressure detection units 4, the bottom acoustic detection component 5, and the four sets of electric louvered guide vanes 31 to perform time synchronization and calculation processing within the same controller. The control component can also acquire existing operating parameters of the air supply component 2, such as the cooling fan speed, main air supply pressure or air supply flow rate, and the current opening degree of the louvered guide vanes, and use these operating parameters as current operating status information to establish a dynamic transport time reference corresponding to normal production status.

[0065] First, the control component identifies abnormal material transport events in the central region of the granulation tower 1 based on dynamic pressure disturbance information.

[0066] Specifically, during real-time production, the control component synchronously acquires dynamic pressure signals from three dynamic pressure detection units 4 according to a unified time base. The control component first performs basic data cleaning on the three pressure detection channels, including outlier removal, mean removal, detrending, and digital filtering. For single-point spikes caused by electromagnetic interference or instantaneous sampling anomalies, median filtering or Hampel outlier detection methods can be used for removal; for slow trends caused by fan operation, tower temperature changes, and the overall chimney effect, moving average or linear detrending methods can be used for removal.

[0067] Furthermore, a Fast Fourier Transform (FFT) is performed on the dynamic pressure signal to determine the main dynamic frequency components contained in the current signal. The control component uses the FFT spectrum to observe the distribution of dynamic energy in the normal operation window and the abnormal pressure event window, and determines the target frequency band that can stably reflect the pressure disturbance when abnormal material passes through the detection area during the prototype calibration stage. In one embodiment, the pressure fluctuation under normal conditions is mainly concentrated in the low-frequency, slowly changing part, while the abnormal material will generate short-term dynamic disturbances in a higher frequency range when passing through the detection area. During the prototype calibration stage, the control component performs FFT analysis on multiple sets of normal pressure signals and pre-generated slightly abnormal transport pressure signals, and selects the passband of the digital bandpass filter based on the frequency band that can be stably distinguished between the two state spectra. The digital filter can be a fourth-order Butterworth bandpass filter or an FIR bandpass filter. The specific upper and lower limits of the passband are determined by the actual tower size, particle settling velocity, pressure tapping chamber structure, and target event duration, and are not fixed to a certain set of frequency values. In one prototype implementation, if FFT analysis confirms that the target dynamic disturbance is mainly located between 10Hz and 100Hz, then bandpass filtering of this frequency band is used in the real-time processing stage. However, this frequency band is only used as a specific embodiment and is not intended to limit the scope.

[0068] The dynamic pressure signals collected by each dynamic pressure detection unit 4 are filtered to establish the dynamic pressure background value corresponding to each detection location. The transient pressure disturbance signal is obtained by calculating the difference. Specifically, the control component further establishes its own dynamic pressure background value for each detection location to avoid differences in normal pressure references caused by installation structure, local airflow, or tower geometry differences among the three locations. For different pressure detection locations, a moving average can be used to obtain their respective dynamic pressure background values. Then, the real-time dynamic pressure signal collected by each dynamic pressure detection unit 4 is subtracted from the corresponding dynamic pressure background value to obtain the transient pressure disturbance signal corresponding to each dynamic pressure detection unit 4.

[0069] Furthermore, the control component extracts abnormal pressure events based on transient pressure disturbance signals. Specifically, to avoid misjudging a single noise spike as an abnormal material event, the control component does not use whether a single sampling point exceeds a threshold as the sole criterion during actual operation. Instead, it considers the disturbance amplitude, duration, and event energy of the transient pressure simultaneously. When the transient pressure disturbance of a certain dynamic pressure detection unit 4 continuously exceeds the event detection threshold determined based on the current dynamic noise background and continues to reach a preset minimum duration, the control component marks this time period as a candidate pressure event. For a candidate pressure event, the control component records the event start time, peak pressure disturbance amplitude, event duration, and the event energy index used to characterize the signal strength of the event.

[0070] When abnormal material particles or particle groups pass through the pressure tapping detection area corresponding to the dynamic pressure detection unit, their movement will cause short-term pressure disturbances in the local gas-solid two-phase flow field, causing the dynamic pressure signal to deviate from the normal pressure background. For transient pressure disturbance signals at the same detection location and under the same sampling conditions, the larger the disturbance amplitude, the greater the degree to which the pressure signal deviates from the normal background within the corresponding time period; at the same time, the longer the disturbance duration, the longer the abnormal pressure state continues to act on the detection area. Therefore, in order to simultaneously reflect the amplitude and duration of the pressure disturbance in the same index, this application adopts the sum of squares in the time domain, commonly used in the field of signal processing, as the signal energy index of candidate pressure events.

[0071] Specifically, the energy index of abnormal events can be calculated using the following existing conventional discrete signal energy formula:

[0072]

[0073] in, This represents the abnormal event energy index of a single candidate pressure event corresponding to the i-th pressure detection location; This represents the transient pressure disturbance value at the nth sampling point within the event window; This indicates the total number of sampling points included in the event window. Since the transient pressure disturbance value has been processed against the dynamic pressure background benchmark before calculation, both its positive and negative disturbances are squared before being accumulated. This avoids the positive and negative pressure disturbances canceling each other out when directly summed. At the same time, the contribution of larger transient pressure disturbances to the signal energy index is relatively increased after squaring, thus enabling the index to reflect both the disturbance amplitude and duration of abnormal pressure events.

[0074] The squared and summed amplitudes of the aforementioned transient pressure disturbances are not interpreted as the actual mechanical energy of the abnormal material particles. Instead, they are quantified using a signal energy metric commonly used in signal processing to measure the dynamic signal intensity of a single abnormal pressure event. Therefore, in this embodiment, event energy is a signal processing indicator used to characterize the strength of abnormal pressure disturbances, not mechanical energy with Joule dimensions. Since all pressure detection events use the same pressure sensor, the same sampling method, and the same signal processing flow, the event energy indicator can be used to compare the disturbance intensity of different abnormal pressure events under the same detection channel and operating conditions.

[0075] For other implementations that require further consideration of the actual dimensions of the event duration, time integration can be performed based on the above discrete signal energy index combined with a unified sampling period. However, in this embodiment, each dynamic pressure detection unit 4 uses a unified sampling frequency, and abnormal pressure events are compared under the same sampling conditions. Therefore, the above discrete signal energy index can be directly used as the event intensity feature. The control component further combines the occurrence time of the abnormal event, the amplitude of the pressure disturbance, the duration, and the event energy index to form an abnormal pressure event feature set, and uses the characteristic changes of multiple consecutive abnormal pressure events to determine the changing trend of the material transport status in the middle of the tower.

[0076] Furthermore, the control component records each abnormal pressure event as an event data set, including at least the event occurrence time, event peak value, event duration, and event energy. Multiple events obtained during continuous operation are then arranged chronologically to form a sequence of abnormal material transport events in the middle of the tower. It should be noted that the abnormal material transport events in this application do not require each pressure event to correspond one-to-one with an independent material particle. Instead, multiple material particles are allowed to pass through the detection area simultaneously, forming a single event. Furthermore, a batch of abnormal materials is allowed to generate multiple events consecutively within a certain time frame. Therefore, subsequent transport time calculations utilize statistical time correlation between event sequences, rather than calculating based on the absolute trajectory of a single particle.

[0077] Furthermore, the acoustic detection component 5 synchronously and continuously acquires acoustic signals from the bottom of the tower. Since normal particle impact noise is continuously generated in the bottom receiving area, the control component first performs bandpass filtering and background noise cancellation on the original acoustic signal, and then extracts transient acoustic events that characterize concentrated material impact from the acoustic signal. For the acoustic signal, FFT or short-time Fourier transform can be used to determine the main frequency band of normal material falling noise and the frequency band corresponding to abnormal impacts, and then a digital bandpass filter is used to retain the target frequency band. When enhanced transient impact detection is required, Hilbert transform can be used on the filtered acoustic signal to obtain an analytical signal, and its envelope can be calculated to obtain the acoustic energy envelope that changes over time. Similarly, when the acoustic envelope exceeds the dynamic background threshold and continues for a set time, a bottom material falling acoustic event is generated. Each acoustic event records its occurrence time, peak value, duration, and acoustic energy, ultimately forming a bottom acoustic event sequence.

[0078] Specifically, to avoid directly mistaking the normal background noise generated by continuous particle falling as an abnormal falling event, acoustic event detection also adopts a dynamic background method. The control component first establishes the bottom acoustic background energy range within the normal production window, and then uses the difference or ratio between the current acoustic envelope and the background energy to determine whether there is a significant impact event. For continuous acoustic events, the control component can use an event segmentation method to merge adjacent but closely spaced impact events into the same event group, so that the acoustic event sequence reflects the time variation of abnormal bottom falling activity, rather than being affected by the collision interval of individual particles.

[0079] Subsequently, time correlation analysis was performed on abnormal material transport events and acoustic event information to determine the apparent transport time of the material moving from the middle area of ​​the granulation tower 1 to the bottom collection area.

[0080] Specifically, after obtaining the abnormal transport event sequence in the middle of the tower and the acoustic event sequence at the bottom, the control components perform cross-height time correlation analysis. This time correlation is not a direct cross-correlation of the original pressure waveform and the original acoustic waveform, because the corresponding physical quantities and waveform shapes are different. Instead, the detected events are first divided into several time windows according to a unified time axis, for example, 100ms or 200ms per window. The total intensity of the abnormal pressure events and the total intensity of the acoustic events at the bottom are then statistically analyzed within each time window, thus forming the middle abnormal event intensity sequence and the bottom acoustic event intensity sequence, respectively.

[0081] Specifically, for the k-th time window, the anomalous event intensity of that time window can be constructed based on the pressure event energy, number of events, and event amplitude within that window. To avoid the influence of differences in the dimensions of different sensors on the cross-correlation results, the event intensity sequences are then normalized, making the two sequences dimensionless relative event intensity sequences. Subsequently, the existing normalized cross-correlation algorithm is used to calculate the correlation between the two event sequences under different time delays. A basic form of normalized cross-correlation is:

[0082]

[0083] in, Indicates a delay of Normalized cross-correlation coefficients for each sampling time window This represents the data for the k-th time window of the intensity sequence of abnormal transport events in the central region. Indicates the intensity sequence of bottom acoustic events in delay The data corresponding to each time window. and These represent the average values ​​of the two sequences within the current association window. This indicates the number of sampling points representing the time window delay. The control component gradually changes within a preset time search range. The system identifies the location where the normalized cross-correlation coefficient reaches its maximum value and uses the time delay corresponding to this location as the apparent transport time of the current abnormal material event from the detection area in the middle of the tower to the collection area at the bottom. Before performing the normalized cross-correlation calculation, the control component first determines whether there are valid changes in the intensity sequences of the abnormal transport events in the middle and the acoustic events at the bottom within the current correlation window. When either event intensity sequence is a constant value within the current window, resulting in zero deviation between the corresponding sequence and its average value, the denominator of the normalized cross-correlation formula is zero. In this case, it is considered that the current correlation window lacks valid dynamic events that can be used to determine the time delay, and the cross-correlation calculation for this window is not performed. The system continues to acquire subsequent correlation windows. Only when both event intensity sequences have valid changes within the current window is the normalized cross-correlation calculation performed and the corresponding time delay determined.

[0084] Let the sampling time window of the event sequence be... The delay time corresponding to the peak cross-correlation is:

[0085]

[0086] in, This indicates the apparent transport time of the current abnormal material event. This represents the number of delay windows required to reach the maximum value of the cross-correlation coefficient. This indicates the duration of each time window in the event sequence. The control component does not... Instead of interpreting it as the precise free fall time of a single particle, it is used as a statistical time characteristic of the propagation of a batch of abnormal material transport events from the central detection area to the bottom of the tower.

[0087] To avoid erroneous correlations, the control component sets a reasonable time search range when performing cross-correlation analysis. The time search range is determined based on the effective detection height of the granulation tower 1, the normal material conveying speed, and historical normal conveying time. For example, if the normal apparent conveying time in a specific production line is approximately 12 seconds, the correlation search range can be set to 5-20 seconds, searching for correlation peaks only within this time interval. If the maximum cross-correlation coefficient is too low, it indicates that there is no sufficiently stable time correlation between the pressure event in the middle of the current window and the acoustic event at the bottom. In this case, the control component does not directly update the apparent conveying time but continues to use the most recent valid conveying status result until a new event window that meets the valid correlation conditions is obtained.

[0088] In actual operation, due to the time lag in bottom acoustic detection within the tower, this embodiment does not treat bottom acoustic events as mandatory triggering conditions for every control action. The control component loads a dynamic transport time reference corresponding to the current operating condition at the start of operation and obtains the latest valid reference after reliably correlating the most recent mid-level event with the bottom acoustic event. When a new abnormal pressure event recurs in the middle of the tower, the control components can directly determine whether to initiate primary or secondary airflow regulation based on the most recently confirmed transport status and the spatial distribution of the current pressure event in the middle, without waiting for the new event to reach the bottom of the tower again. Once the batch of material finally reaches the bottom of the tower and generates a new acoustic event, the system will recalculate through cross-height event correlation. This is used to calibrate the transport state model. Therefore, in this embodiment, the bottom acoustic detection mainly serves to confirm cross-height events and calibrate the transport state, while the middle pressure detection serves to quickly detect anomalies and provide real-time control input.

[0089] After obtaining the currently effective apparent transport time Subsequently, the control component compares it with a pre-established and online-maintained dynamic transport time reference. In this embodiment, the dynamic transport time reference is not randomly generated temporarily after each anomaly is detected, nor is it directly updated using the current anomaly data when an anomaly is detected. Instead, it is initially established before actual production begins and slowly updated using only data that has been determined to be in a normal state during production, in order to avoid anomaly data being mistakenly learned as a normal state. Specifically, during the prototype debugging and the first operation of the production line, several normal production conditions that can represent the actual production process are first selected, such as different cooling air flow rates, different fertilizer material flow rates, different tower internal temperature states, and different initial louver opening states. The granulation tower 1 is run under these conditions, and after confirming that the particle size, moisture content, temperature, and bottom material collection state are all within the normal production range, the abnormal event sequence in the middle of the tower and the acoustic event sequence at the bottom are collected over a period of time to obtain multiple sets of normal apparent transport time data.

[0090] For example, for each normal operating condition, the median or moving average can be used as the initial normal transport time to reduce the impact of outliers caused by occasional shocks and events. Normal data can be divided into multiple operating condition intervals based on cooling air flow, material flow, tower temperature, and louver opening. This allows the control components to select the corresponding normal dynamic transport time benchmark based on the current air supply, material flow, temperature, and louver opening during actual operation, rather than using a fixed value for all production states. For instance, when the current airflow is high, the benchmark established under high airflow conditions can be selected; when the current airflow is low, the benchmark established under low airflow conditions can be selected. For actual states between two calibrated operating conditions, interpolation can be performed using the benchmark values ​​of adjacent normal operating conditions or using the normal benchmark value closest to the current operating condition, thus allowing the dynamic transport benchmark to change with the operating condition.

[0091] After establishing the initial normal dynamic transport time benchmark, the system enters the continuous production phase. The control component only updates the dynamic transport time benchmark slowly when the current production state is judged to be normal. For example, the original normal benchmark is allowed to be corrected using the newly obtained apparent transport time only when the following conditions are met simultaneously in the current window: the apparent transport time is within the normal allowable range, the spatial difference of the three dynamic pressure detection units 4 does not exceed the normal range, the bottom acoustic event does not show persistent abnormality, and no abnormal control state above level one has been entered. In one embodiment, the dynamic transport benchmark is slowly updated using an exponentially weighted moving average. Specifically, the normal dynamic transport time benchmark can slowly follow the normal transport changes caused by seasonal temperature, material formulation, and equipment aging, without significant jumps due to a single measurement result. When the control component judges that the current operating condition has entered an abnormal state, the updating of the dynamic transport time benchmark is paused, and the benchmark update is resumed only after multiple consecutive detection windows have returned to normal.

[0092] In some implementations, to avoid inconsistencies in judgment criteria between different operating conditions caused by simply using time differences, the relative deviation of apparent transport time can be calculated:

[0093]

[0094] in, Indicates the relative degree of anomaly in apparent transport time. Indicates the dynamic transport time base. This indicates the apparent transport time of the current abnormal material event. The control component does not rely solely on a single event. Instead of immediately classifying deviations as severe anomalies, the system combines the direction and magnitude of deviations across multiple consecutive associated windows with the frequency of abnormal events to determine the trend of material transport status changes. Based on the results, the anomaly level is categorized into three control judgment levels, from minor to severe. For example, if the current apparent transport time deviates slightly from the dynamic baseline only once and subsequently returns to the normal range, only one short-term deviation is recorded, and the control level remains unchanged. If deviations in the same direction occur across multiple consecutive effective windows, and the deviation magnitude gradually increases, it is determined that the current material transport status is continuously deviating from the normal state, and the system enters the first-level trend suppression control judgment. If the deviation reaches the severe anomaly range and there is a significant increase in central pressure disturbance and bottom acoustic impact, the system enters the third-level protection control judgment.

[0095] Finally, based on the apparent transport time characterizing the material transport state and the circumferential pressure difference information among multiple dynamic pressure detection units 4, the airflow regulating component 3 is controlled to adjust the airflow distribution inside the granulation tower 1.

[0096] Specifically, the control component determines the circumferential distribution of abnormal transport states based on the abnormal pressure event characteristics acquired by the three dynamic pressure detection units 4. The number, peak value, duration, and energy of corresponding abnormal events are obtained at each of the three detection locations, and these characteristics are converted into pressure disturbance intensity indices for the three circumferential detection areas. To avoid misjudgment of spatial state due to a single abnormal impact, the spatial distribution judgment uses statistical results within at least one continuous time window, rather than the instantaneous peak value at a single sampling point. The control component can calculate the average energy, number of events per unit time, and peak distribution of abnormal events in each region within the current window, and then use these parameters according to pre-defined weights to form the pressure disturbance characteristic parameters for the corresponding region.

[0097] The specific logic for judging the circumferential pressure difference information within the tower is as follows: when the pressure disturbance intensity indices of the three circumferential detection areas are at similar levels and show no significant changes over multiple consecutive windows, the abnormal material transport is considered to be relatively uniformly distributed circumferentially within the tower. When the intensity of the abnormal event in one area is consistently significantly higher than that in the other two areas, it is considered that the abnormal material transport tends to concentrate in that circumferential area. When two adjacent areas are simultaneously higher than the remaining areas, the abnormal area is considered to be located between the two detection positions. For the degree of spatial non-uniformity, the standard deviation and mean of the pressure disturbance intensity in the three areas can be used to calculate the dispersion. Both the standard deviation and mean are calculated using existing statistical methods. When the mean exceeds a preset minimum effective value, the ratio of the standard deviation to the mean can be used as a dimensionless spatial dispersion index. When the mean is close to zero, the maximum difference between the three detection points is directly used as the basis for judging spatial anomalies to avoid instability caused by dividing by a value close to zero.

[0098] Furthermore, after establishing the degree of apparent transport time anomalies and their circumferential distribution, the control unit determines the airflow regulation strategy at the bottom of the tower. Airflow regulation does not presuppose that all anomalies will be addressed by simply increasing the cooling airflow. Instead, it determines the regulation direction based on the deviation of the current transport state from the dynamic normal baseline and the pressure status of the three circumferential detection zones. If the current apparent transport time is consistently lower than the dynamic transport time baseline, and prototype calibration has confirmed that increasing the reverse airflow can move the material's apparent transport time back to the normal range, the control unit increases the cooling airflow in the corresponding zone. If actual equipment calibration results indicate that the current state requires a reduction in airflow to restore normal transport time, the control unit executes flow regulation in the opposite direction. Therefore, the airflow regulation direction is not arbitrarily fixed to increase airflow for anomalies, but is determined by the actual process response relationship of the unit.

[0099] In some implementations, to enable the three pressure detection zones to control the four sets of louvered guide elements 31, the influence relationship between each set of louvered guide elements 31 and the three pressure detection positions is first established during the prototype installation and commissioning phase. Under normal operating conditions, the four sets of louvered guide elements 31 are subjected to small-amplitude independent opening disturbances. For example, a preset opening increment is added to the first set of louvered guide elements 31 under stable operating conditions, while keeping the openings of the other three sets of louvered guide elements 31 unchanged. At the same time, the stable changes at the three pressure detection points are recorded. Then, the original opening of the first set of louvered guide elements 31 is restored, and the same test is performed on the second, third, and fourth sets of louvered guide elements 31 in sequence. Each test is repeated multiple times, and the average response is taken to reduce the impact of fan fluctuations and environmental noise. Finally, the local pressure response relationship of the four sets of actuators to the three pressure detection points is formed. Through the above calibration, the control component can obtain the influence relationship table between the opening changes of each set of louvered guide elements 31 and the three pressure detection points, and adjust the airflow according to the three levels of protection control judgment.

[0100] For example, in the first-level trend suppression control, when the apparent transport time deviates continuously from the dynamic transport time reference, but the three pressure detection areas have not yet shown severe spatial inhomogeneity, the control component first performs a small-scale overall airflow adjustment. At this time, the four sets of louvered guides 31 can make small opening changes in the same direction, increasing or decreasing the intensity of the overall cooling airflow, and maintaining the current opening for a preset flow field response hold time. After the hold time ends, the three dynamic pressure channels are re-detected to determine whether the intensity of the abnormal pressure event and the spatial distribution inside the tower have changed in the normal direction. To avoid continuous rapid back-and-forth movements of the servo actuator, a minimum hold time is set between two adjacent overall airflow adjustments, and the amount of change in louver opening in a single instance is limited.

[0101] In the second-level spatial correction control, when the three dynamic pressure detection units 4 show that abnormal pressure disturbances are significantly concentrated in a certain circumferential region, the control component no longer uses the four sets of louvered guides 31 to change synchronously in the same direction. Instead, it uses the influence relationship obtained from the aforementioned calibration to calculate the coordinated opening of multiple louvered guides 31. For example, when the pressure disturbance characteristics of the first pressure detection area are consistently high, while the pressure disturbance characteristics of the third pressure detection area are low, the control component seeks a louver combination that can reduce the dynamic pressure disturbance in the first detection area and improve the response in the third detection area, while trying to keep the second detection area within the target range. Since the four sets of louvered guides 31 have one more degree of freedom than the three pressure detection quantities, the fourth degree of freedom can be used to simultaneously adjust the overall airflow intensity or keep the total air supply basically unchanged, thereby correcting the circumferential flow deviation while avoiding unnecessary large changes in the overall cooling capacity.

[0102] In some implementations, the control component uses an iterative solution method with upper and lower limit constraints to determine the target opening of the four sets of louvered guide elements 31. The control component first determines the required pressure correction direction based on the deviation between the current pressure value and the target pressure value at three pressure detection points in the middle of the granulation tower 1. Then, using a pre-calibrated influence relationship table, it calculates multiple candidate louvered guide element 31 opening combinations. While satisfying the minimum opening, maximum opening, maximum single opening change, and total fan air supply capacity limits for each set of louvered guide elements 31, it selects the opening combination that reduces the total error at the three pressure detection points. After the action is completed, the dynamic pressure signals corresponding to the three dynamic pressure detection units 4 are remeasured. If the deviation at the three pressure measurement points decreases as expected, the control direction is maintained or further adjustments are made in smaller steps; if the deviation increases instead, the adjustment direction is corrected in the opposite direction based on the latest pressure response. This eliminates the need to pre-assume that a higher pressure on one side necessarily means the corresponding louver guide 31 will be closed, and instead the louver guide 31 on the opposite side will be enlarged. Instead, the correspondence between different louver guides 31 and different pressure detection areas is determined through actual device calibration data.

[0103] In the third-level severe anomaly protection control, when the apparent transport time deviates significantly from the dynamic transport time reference, and the amplitude, duration, or energy of the mid-section pressure disturbance event increases significantly, and the bottom acoustic detection component 5 confirms a significant abnormal material drop acoustic event, the control component enters the enhanced cooling protection mode. At this time, the control component increases the overall cooling airflow to the preset protection range, while maintaining the air supply difference between each circumferential region within a safe set value to reduce the risk of abnormal material that has not yet been cooled or solidified continuing to concentrate at the bottom of the tower. The specific maximum airflow, maximum louver opening, and holding time of this protection mode are determined during the equipment commissioning phase through thermal balance and tower pressure tests, and are constrained by the fan capacity, the tower's allowable pressure, and the process boundary preventing material from being blown back.

[0104] In the aforementioned three-level protection control, the role of bottom acoustic events is primarily to confirm that a severe anomaly has occurred, rather than serving as the sole trigger condition for action. When the dynamic pressure detection data in the middle of the tower indicates that the abnormal transport state is continuously deteriorating, and the most recent effective apparent transport time is still within the effective time window of the control component, the control component can directly initiate protection actions based on the detection results in the middle of the tower, without waiting for a new bottom acoustic impact event to occur. The bottom acoustic information is then used to confirm the degree of anomaly before control and to readjust the apparent transport time, thereby avoiding the overall delay in control action caused by the inherent time lag of bottom detection.

[0105] After the airflow adjustment component 3 completes a control action, the control component does not directly consider the control to be complete, but instead performs feedback verification. First, it re-acquires the pressure disturbance signals from the three dynamic pressure detection units 4, and obtains a new pressure event sequence according to the aforementioned detrending, spectrum analysis, digital filtering, and abnormal event extraction process. Then, it continues to acquire the acoustic signals from the acoustic detection component 5. When the subsequent material reaches the bottom collection area and forms a sufficiently stable acoustic event correlation, the apparent transport time is recalculated. The control component compares the apparent transport time, abnormal event frequency, pressure disturbance intensity, and spatial differences between the three detection areas before and after the adjustment to determine whether the airflow adjustment is effective.

[0106] If the apparent transport time recovers towards the dynamic transport time reference direction after adjustment, and the difference between the three dynamic pressure detection zones decreases, while the duration and energy of the abnormal event decrease, then the control is considered effective. The control component reduces the subsequent control step size or gradually restores the four sets of louver guides 31 to their normal operating opening. If the abnormal transport state does not significantly weaken after adjustment, the current control level is maintained and the next small step adjustment is performed; if the abnormality worsens further, the control level is increased. If the control component detects that a certain louver guide 31 has reached the limit of permissible action, but the pressure space difference and the degree of transport abnormality have not been improved, the adjustment range is stopped from being expanded further, and an equipment abnormality prompt is output so that the operator can check the status of the fan, air duct, louver mechanism, or granulation process.

[0107] In this embodiment, the dynamic transport time reference is not updated under abnormal conditions. For example, when the current apparent transport time continues to shorten and the pressure disturbance energy continues to increase, the control component freezes. Abnormal data is not allowed to participate in baseline updates; the exponentially weighted update is only restarted when the material transport status returns to normal and continues to meet the normal status determination conditions. This avoids the dynamic baseline gradually drifting with the abnormal state due to long-term abnormal operation, causing the controller to mistakenly identify the abnormal state as a new normal production state.

[0108] In actual continuous production, the control components operate simultaneously on two time scales. On one hand, signals from the three dynamic pressure detection units 4 are continuously processed over a shorter time period to quickly detect new abnormal transport events and determine the circumferential distribution status. On the other hand, acoustic events are continuously processed over a longer time scale, which is required for the material to move from the central detection area to the bottom collection area, to confirm the temporal correlation between the aforementioned central events and the bottom material drop and to update the apparent transport time. In this way, the dynamic pressure detection components constitute a rapid process sensing channel, and the acoustic detection components 5 constitute a cross-height event confirmation channel. The two establish a unified time scale through the apparent transport time, and the control components then use this time scale to determine whether the subsequent material is still in an interventionable stage and whether the current airflow adjustment needs to be executed in advance.

[0109] This application does not directly interpret apparent transport time as the aerodynamic equivalent diameter of a certain particle size or agglomerate. Instead, it uses apparent transport time as a process characteristic parameter that comprehensively reflects particle settling, reverse airflow, material load, and changes in local airflow state. Since transport time varies with air supply, material flow rate, temperature, and louver opening under normal production conditions, a pre-established and slowly updated dynamic transport time benchmark allows for comparison of current detection results with normal conditions under similar or near-normal production conditions. This reduces the impact of environmental and process changes on anomaly detection.

[0110] This embodiment of the application sets up a dynamic pressure detection component and an acoustic detection component 5 distributed circumferentially in the middle of the tower body during the high-tower granulation process. This allows the control component to first identify abnormal material transport events using dynamic pressure disturbance information in the middle of the tower body, then obtain the apparent transport time using the time correlation between the acoustic events at the bottom of the tower and the abnormal events in the middle, and judge the degree of abnormal material transport by combining the dynamic transport time benchmark. At the same time, the spatial distribution of the abnormal state is judged by the pressure disturbance difference between the three circumferential detection positions. Then, the cooling airflow in different areas is adjusted as a whole or asymmetrically by four sets of independently controlled airflow adjustment components 3. The adjustment effect is verified and corrected by the pressure disturbance detected again and the subsequent apparent transport time, thus forming a complete closed-loop control process. This improves the process perception capability and airflow adjustment targeting of abnormal material transport states during the high-tower granulation process without directly contacting the high-temperature moving materials inside the tower.

[0111] In the description of this application, the terms "first" and "second" are used 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" or "second" may explicitly or implicitly include one or more features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0112] The embodiments, implementation methods, and related technical features of this application can be combined and substituted for each other without conflict.

[0113] The above are merely preferred embodiments of this application and are not intended to limit this application in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application without departing from the scope of the technical solution of this application shall still fall within the scope of the technical solution of this application.

Claims

1. A high-tower granulation functional fertilizer production device, comprising a granulation tower body (1), an air supply component (2) for introducing cooling airflow into the granulation tower body (1), and an airflow regulating component (3) for regulating the airflow distribution inside the granulation tower body (1), wherein the granulation tower body (1) is used for granulating a fertilizer mixture composed of multiple fertilizer components and additives, characterized in that, Also includes: The dynamic pressure detection component is located in the middle of the granulation tower body (1) and includes multiple dynamic pressure detection units (4) arranged circumferentially along the granulation tower body (1) to obtain dynamic pressure disturbance information generated by the gas-solid two-phase motion inside the granulation tower body (1) at different circumferential positions. An acoustic detection component (5) is set in the bottom collection area of ​​the granulation tower (1) to acquire acoustic event information generated when the material reaches the bottom collection area; The control component is electrically connected to the dynamic pressure detection component, the acoustic detection component (5), and the airflow regulation component (3), respectively. The control component is configured as follows: Abnormal material transport events in the central region of the granulation tower (1) are identified based on dynamic pressure disturbance information; The time correlation analysis of abnormal material transport events and acoustic event information was performed to determine the apparent transport time of the material from the middle area of ​​the granulation tower (1) to the bottom collection area. Based on the apparent transport time characterizing the material transport state and the circumferential pressure difference information between multiple dynamic pressure detection units (4), the airflow regulating component (3) is controlled to adjust the airflow distribution inside the granulation tower (1).

2. The high-tower granulation functional fertilizer production device according to claim 1, characterized in that, The dynamic pressure detection component includes three dynamic pressure detection units (4). The three dynamic pressure detection units (4) are set at the same height position of the granulation tower (1) and are evenly spaced along the circumference of the granulation tower (1) to obtain dynamic pressure disturbance information in different circumferential areas of the granulation tower (1).

3. The high-tower granulation functional fertilizer production device according to claim 1, characterized in that, The dynamic pressure detection unit (4) includes: The pressure tapping structure (41) is provided on the tower wall of the granulation tower body (1), including a measuring pressure tapping part (411) communicating with the interior of the granulation tower body (1), a protective part (412) for preventing material from entering, and a measuring pressure chamber communicating with the measuring pressure tapping part (411) for acquiring the local dynamic pressure disturbance generated when the material passes through the corresponding detection area; The reference pressure tapping assembly (42) is circumferentially spaced at the same or similar height as the pressure tapping structure (41). The reference pressure tapping assembly (42) includes a reference pressure tapping port (421), a damping connecting pipe (422) communicating with the reference pressure tapping port (421), and a reference pressure chamber (423) communicating with the damping connecting pipe (422), which is used to obtain the slowly changing reference pressure in the tower. The differential pressure detection element (43) is connected to the measuring pressure chamber and the reference pressure chamber (423) respectively, and is used to detect the dynamic pressure difference between the measuring pressure chamber and the reference pressure chamber (423); The signal acquisition module is electrically connected to the differential pressure detection element (43) and is used to acquire the dynamic pressure signal formed by the dynamic pressure difference.

4. The high-tower granulation functional fertilizer production device according to claim 1, characterized in that, The control component is configured as follows: Acquire the dynamic pressure signals collected by each of the dynamic pressure detection units (4); Establish dynamic pressure background values ​​corresponding to each detection location; Based on the difference between the dynamic pressure signal and the dynamic pressure background value, a transient pressure disturbance signal that removes the influence of environmental pressure changes is obtained. Abnormal pressure events are extracted based on the transient pressure disturbance signal.

5. The high-tower granulation functional fertilizer production device according to claim 4, characterized in that, The control component is further configured as follows: The transient pressure disturbance signal is subjected to event feature extraction to obtain the occurrence time, disturbance amplitude, duration, and event energy parameters of the abnormal pressure event; And based on the multiple abnormal pressure events, a sequence of abnormal transport events in the middle of the tower body is formed.

6. The high-tower granulation functional fertilizer production device according to claim 5, characterized in that, The control component is also configured to: Acquire the bottom acoustic signal collected by the acoustic detection component (5); Event recognition is performed on the bottom acoustic signal to obtain the material falling acoustic event sequence; Cross-correlation analysis was performed on the abnormal transport event sequence in the middle of the tower and the acoustic event sequence of material falling, and the time delay corresponding to the peak value of the cross-correlation was taken as the apparent transport time.

7. The high-tower granulation functional fertilizer production device according to claim 1, characterized in that, The control component is also configured to: Acquire historical apparent transport time data under normal production conditions and establish a dynamic transport time benchmark. The currently obtained apparent transport time is compared with the dynamic transport time benchmark to determine the degree of anomaly in the current material transport.

8. The high-tower granulation functional fertilizer production device according to claim 7, characterized in that, The control component is also configured to: The pressure disturbance characteristic parameters corresponding to the multiple dynamic pressure detection units (4) are obtained respectively; Based on the differences between the pressure disturbance characteristic parameters corresponding to different circumferential positions, the distribution state of abnormal material transport status in the circumferential direction of the granulation tower body (1) is determined, so as to control the airflow regulating component (3) at the corresponding position to perform the regulating action.

9. The high-tower granulation functional fertilizer production device according to claim 8, characterized in that, The airflow regulating component (3) includes multiple independently controlled airflow regulating units; The control component is configured to determine the working state of each airflow adjustment unit and execute corresponding adjustment actions based on the comparison result between the apparent transport time and the dynamic transport time reference, as well as the difference between the circumferential pressure disturbance characteristic parameters corresponding to the plurality of dynamic pressure detection units (4).

10. The high-tower granulation functional fertilizer production device according to claim 9, characterized in that, The control component is also configured to: After the airflow regulating component (3) performs the regulating action, the dynamic pressure disturbance information and the acoustic event information are reacquired; The apparent transport time and the circumferential pressure distribution state are re-determined based on the re-acquired dynamic pressure disturbance information and acoustic event information. Based on the changes in material transport status before and after adjustment, the adjustment action of the airflow adjustment component (3) is corrected.

Images