A method for continuous crystallization of a dynamically scheduled water-soluble fertilizer

Abstract

The application provides a kind of dynamically scheduled water-soluble fertilizer crystallization method, comprising: establishing the mapping relationship of logical scheduling unit and physical execution component, and configuring scheduling parameter;The internal space of crystallizer is divided into multiple functional partitions with hierarchical order, and the material state information of each functional partition is packaged as a task object and stored in the corresponding pending task queue;Drive logical scheduling unit to determine the basic resource quota required for distribution steady state based on hierarchical order;Instruction physical execution component executes load output according to basic resource quota;Real-time analysis of the load attribute of the task object, calculate the local crystallization load of each functional partition, and determine the adjustable resource margin according to the difference between the total rated power of the physical execution component and the allocated quota;Real-time monitoring of the task object in the underlying functional partition, when the preset output confidence threshold is met, dynamically adjust the opening of the discharge mechanism based on the target capacity.

Description

Technical Field

[0001] This invention relates to the field of inorganic salt crystallization production technology, and more specifically, to a method for continuous crystallization of water-soluble fertilizers with dynamic scheduling. Background Technology

[0002] Crystallization is a core technology for purifying and separating water-soluble fertilizers in industrial production. Currently, the crystallization production of water-soluble fertilizers generally adopts an intermittent crystallization method. This method requires batch-wise completion of processes such as feeding, cooling, and discharging. The feeding and discharging auxiliary links take a long time, resulting in low equipment utilization, low production capacity, and high energy consumption. At the same time, the manual operation intensity is high, the quality difference between batches of crystallized products is large, the crystal particle size distribution is uneven, and the product is prone to caking.

[0003] To address the drawbacks of intermittent crystallization, various continuous crystallizers and crystallization methods have emerged in existing technologies. However, most continuous crystallization devices are structurally complex, difficult to maintain and repair, and have poor adaptability, making them unsuitable for large-scale water-soluble fertilizer crystallization production. Some continuous crystallization technologies employ methods such as vacuum cooling and forced high-flow-rate circulation, which suffer from high energy consumption and low heat exchange efficiency. Furthermore, uneven heat exchange, crystal sedimentation, and excessive precipitation of fine crystals are prone to occur during the crystallization process, resulting in poor stability of crystal product quality and an inability to simultaneously achieve a balance between production efficiency, energy consumption control, and product quality.

[0004] Therefore, the existing technology has problems, and there is an urgent need for a water-soluble fertilizer continuous crystallization method that is simple in structure, easy to operate, can realize large-scale continuous production, and can significantly reduce energy consumption while ensuring stable crystal product quality and uniform particle size. Summary of the Invention

[0005] (I) Purpose of the invention: In order to solve the problems existing in the prior art, the purpose of this invention is to provide a dynamic scheduling method for continuous crystallization of water-soluble fertilizers, which solves the technical problems of high energy consumption, low production efficiency, difficult equipment maintenance, easy crystal sedimentation, uneven heat exchange, and unstable product quality in the existing water-soluble fertilizer crystallization production, and realizes low energy consumption, large-scale, and continuous crystallization production of water-soluble fertilizers, while ensuring uniform particle size and stable quality of crystal products.

[0006] (II) Technical Solution: To solve the above-mentioned technical problems, this technical solution provides a dynamically scheduled water-soluble fertilizer crystallization method, applied to a continuous crystallization device, the method comprising the following steps: Establish a mapping relationship between logical scheduling units and physical execution components in the controller, and configure scheduling parameters that characterize the weight of resource quota distribution; The internal space of the crystallizer is divided into multiple functional zones with a hierarchical order, and the material status information of each functional zone is encapsulated as a task object and stored in the corresponding task queue to be processed. The logic scheduling unit is driven to determine the basic resource quota required to maintain a steady-state material distribution field inside the crystallizer based on the hierarchical order of each functional partition; the steady-state material distribution field is defined by the basic characteristic attributes of each task object; and the physical execution unit is instructed to perform load output according to the basic resource quota. The load attributes of the task objects in each queue of tasks to be processed are analyzed in real time, the local crystallization load of each functional partition is calculated, and the schedulable resource margin of the current time window is determined based on the difference between the total rated power of the physical execution components and the allocated quota. The system monitors the dequeue readiness status of task objects within the underlying functional partitions in real time. When the dequeue readiness status meets the preset output confidence threshold, the opening of the discharge mechanism is dynamically adjusted based on the target capacity to ensure controlled dequeueing of the corresponding materials.

[0007] Furthermore, the aforementioned dynamic scheduling method for water-soluble fertilizer crystallization also includes, if the schedulable resource surplus is greater than a preset threshold, the logical scheduling unit dynamically calculates the intra-layer weight based on the local crystallization load of each queue of tasks to be processed, and allocates the schedulable resource surplus to the corresponding functional partition according to the intra-layer weight ratio, so as to eliminate task backlog in the corresponding partition.

[0008] Furthermore, the aforementioned method for dynamically scheduling water-soluble fertilizer crystallization also includes converting additional schedulable resource quotas into the instantaneous gain power of the stirring unit or the medium flow rate increment of the cooling unit in the physical execution component; wherein the instantaneous gain power is limited by the total rated power of the physical execution component.

[0009] In addition, the dynamic scheduling method for water-soluble fertilizer crystallization also includes a logic scheduling unit that monitors the changes in material status information of each functional zone after additional allocation in real time. When the local crystallization load returns to the steady-state field range, the logic scheduling unit recovers the instantaneous gain power and returns this part of the resources to the schedulable resource surplus pool.

[0010] Furthermore, the aforementioned dynamic scheduling method for water-soluble fertilizer crystallization also includes the logical scheduling unit extracting the real-time solid content and real-time fluid viscosity from the load attributes of each task object, and calculating the energy demand deviation caused by the solute precipitation rate within each functional zone using a preset phase change resistance model.

[0011] Specifically, the preset phase transformation resistance model includes: establishing a heat accumulation factor within the functional zone due to the heat exchange rate lagging behind the crystal growth rate, and using the heat accumulation factor as a weighted correction parameter for the local crystallization load.

[0012] Specifically, the method for obtaining the material state information includes: setting pressure sensing points and power consumption monitoring points in each of the functional zones, calculating the real-time solid content using static pressure difference, and mapping the real-time fluid viscosity using the operating current fluctuation of the physical execution component.

[0013] The determination of the dequeue readiness state includes: real-time calculation of the local apparent density of materials of task objects within the underlying functional partition; when the apparent density is within a preset finished product density range, the dequeue readiness state is determined to have reached the output confidence threshold.

[0014] The opening adjustment strategy of the discharge mechanism includes: establishing a time correlation function between the target capacity and the discharge readiness index, and using pulse width modulation logic to dynamically adjust the opening window of the discharge mechanism so that the total amount of material discharged per unit time matches the preset target capacity.

[0015] The controlled discharge of the corresponding materials also includes monitoring the steady-state fluctuations of the material distribution in the bottom functional area during the discharge process. If the discharge causes an abnormal increase in the local crystallization load of the area, the logic scheduling unit will trigger emergency resource scheduling to maintain production continuity by increasing the material replenishment rate from the upper functional area to the lower area.

[0016] The functional zones are divided into a core mixing zone, annular circulation zone, and bottom enrichment zone based on the physical location of the internal guide tube of the crystallizer.

[0017] The departure determination and emission control also include synchronization with the downstream centrifugal separation system: The logic scheduling unit obtains feedback on the real-time processing frequency and filter cake moisture content of the downstream centrifuge; when the centrifuge is in a standby cycle and the dequeue ready state meets the threshold, the discharge mechanism is triggered to perform pulse-type start; if the centrifuge reports an overload signal, the logic scheduling unit forces the task object to be suspended in the queue of pending tasks by lowering the output confidence threshold of the dequeue ready state.

[0018] The opening adjustment of the discharge mechanism is based on the correlation curve between historical discharge flow rate and local apparent density to predict the amount of disturbance to the steady field of the underlying functional area during the discharge process. At the same time as the discharge mechanism is opened, the material conveying power from the core mixing zone to the bottom layer is increased in a compensatory manner to offset the negative pressure fluctuation caused by the discharge.

[0019] The mapping logic for the real-time fluid viscosity includes: The effective power output of the stirring unit at different frequencies is collected, and combined with the material density obtained from the pressure sensing point in the current partition, the dynamic viscosity of the task object is obtained by reverse calculation through the correlation model of fluid flow field Reynolds number and power criterion number.

[0020] In the aforementioned dynamic scheduling method for water-soluble fertilizer crystallization, when applied to the production of calcium nitrate tetrahydrate, the hierarchical functional partitions are configured with specific temperature gradients. The cooling rate gradient from the core mixing zone to the bottom enrichment zone is set to 0.5 degrees Celsius to 1.2 degrees Celsius per hour; when the local crystallization load shows a cooling rate deviation exceeding ±10%, the available schedulable resources are triggered to perform gain allocation to the cooling unit.

[0021] Among them, the viscosity threshold of the calcium nitrate tetrahydrate task object is set as follows: When the real-time fluid viscosity of the target object exceeds 150 mPa·s, the logic scheduling unit determines that a local crystallization and agglomeration trend has occurred, and forcibly increases the local shear force output of the stirring unit until the viscosity returns to the steady-state range of 80 to 100 mPa·s.

[0022] (III) Beneficial Effects: This invention provides a dynamic scheduling method for continuous crystallization of water-soluble fertilizers. By dividing the internal space into multiple functional zones and encapsulating the material state of each zone as an independent task object, it can perceive and respond to dynamic changes in local areas within the crystallizer with finer segmentation. By maintaining a steady-state field of material distribution defined by the basic attributes of each zone and driving the physical execution components to operate according to the calculated basic resource quota, it can actively construct and maintain an ideal crystallization environment. This effectively suppresses problems such as local oversaturation, uneven temperature, or poor mixing, thereby significantly improving the uniformity of crystal particle size distribution and ensuring the stability of product quality within and between batches. By calculating the local crystallization load of each zone in real time and combining it with the schedulable resource margin of the physical equipment, limited resources are dynamically and preferentially allocated, avoiding the average distribution and waste of resources, and achieving optimal energy efficiency while ensuring process effectiveness. By monitoring the dequeue readiness status of the task objects in the lower functional zones and dynamically adjusting the discharge mechanism when the output confidence threshold is reached, it achieves linkage control between output and quality. This allows production capacity to closely follow the actual completion of the crystallization process, enabling smooth and adaptive adjustments and avoiding blind material output or production capacity bottlenecks. Attached Figure Description

[0023] Figure 1 This is a flowchart of the steps of a dynamically scheduled continuous crystallization method for water-soluble fertilizers provided by the present invention; Figure 2This is a schematic diagram of the crystallizer provided by the present invention.

[0024] Reference numerals in the attached drawings: 1-body, 2-top cover, 3-external insulation layer, 4-drive motor, 5-stirring shaft, 6-stirring blade, 7-external heat exchanger, 8-circulating pump, 9-frequency converter, 10-controller, 11-first feed inlet, 12-second feed inlet, 13-discharge outlet, 14-level gauge, 15-temperature probe, 16-ultrasonic probe, 17-mother liquor pump, 18-mother liquor tank. Detailed Implementation

[0025] The present invention will be further described in detail below with reference to preferred embodiments. More details are set forth in the following description in order to provide a full understanding of the present invention. However, the present invention can obviously be implemented in many other ways different from those described herein. Those skilled in the art can make similar extensions and derivations based on actual application situations without departing from the spirit of the present invention. Therefore, the scope of protection of the present invention should not be limited by the content of this specific embodiment.

[0026] The accompanying drawings are schematic diagrams of embodiments of the present invention. It should be noted that these drawings are for illustrative purposes only and are not drawn to scale, and should not be construed as limiting the actual scope of protection of the present invention.

[0027] This embodiment provides a dynamically scheduled continuous crystallization method for water-soluble fertilizers. The method is based on a controller installed in a continuous crystallization system. The system includes a mixer, a crystallizer, a centrifuge, and a mother liquor tank. The mixer is connected to a controller or similar device. The crystallizer has an inlet and an outlet, and the mixer is connected to the inlet of the crystallizer. The product solution enters the mixer through an inlet pipe. The inlet of the centrifuge is connected to the outlet of the crystallizer, and the centrifugal rotation of the centrifuge is used to separate the slurry discharged from the bottom outlet of the crystallizer into solid and liquid components. The mother liquor tank is connected to the liquid outlet of the centrifuge and is connected to the mixer through a circulation pipe. Figure 1 The method shown includes the following steps.

[0028] Step 1: Establish the mapping relationship between logical scheduling units and physical execution units in the controller, and configure scheduling parameters that represent the weight of resource quota distribution.

[0029] The aforementioned mapping relationship is established in the controller, where the scheduled resource quotas correspond to the executed load outputs. Scheduling parameters are configured, representing the resource quota distribution weights. For example, 10% is 0.1, 60% is 0.6, etc. These scheduling parameters correspond to different functional zones; each functional zone has a corresponding parameter used to determine the resource quotas. Based on these parameters and the mapping relationship, the logical scheduling unit schedules the corresponding resource quotas and allocates these quotas, i.e., the load output, through the physical execution unit.

[0030] This method includes two scheduling phases. The first scheduling phase, namely the basic supply quota, involves the controller allocating a basic cooling or heating resource quota based on the hierarchical weight of each region, i.e., the preset basic process requirements. For example, 60% of the total steam volume and cooling medium is reserved as a basic quota to ensure the basic stability of the temperature gradient within the 120 m³ cubic meter system and to maintain the minimum effective stirring speed, such as the bottom line of the range of 50~120 r / min, to ensure that the crystals do not settle.

[0031] The volume of the crystallizer mentioned above is 120m³. 3 In high-intensity continuous crystallization production with a daily output of 120 tons, the first stage of basic quota scheduling is implemented. At this time, the 37kW stirring system, which includes an agitator, is allocated 22kW as the basic power to maintain the suspension of the entire 120m³ of material in the tank.

[0032] Step 2: Divide the internal space of the crystallizer into multiple functional zones with hierarchical order, and encapsulate the material status information of each functional zone into task objects and store them in the corresponding task queue.

[0033] The functional zones are divided into a core mixing zone, annular circulation zone, and bottom enrichment zone based on the physical location of the internal guide tube of the crystallizer.

[0034] The method for obtaining the material state information includes setting pressure sensing points and power consumption monitoring points in each of the functional zones, calculating the real-time solid content using static pressure difference, and mapping the real-time fluid viscosity using the operating current fluctuation of the physical actuator.

[0035] Dual pressure sensing points with a spacing of are set at the vertical height of the functional area, including pressure sensor 1 and pressure sensor 2. The apparent density of the slurry is calculated in real time using the static pressure difference formula, and then the real-time solid content is calculated online based on the density difference between calcium nitrate crystals and mother liquor.

[0036] Static pressure difference formula ΔP=ρgh Where ρ is the apparent density of the slurry calculated in real time, g is the gravitational acceleration, and h is the difference between the liquid level and the height.

[0037] The output current and speed of the drive motor are monitored in real time. After eliminating no-load losses using the power characteristic curve, the current real-time fluid viscosity of the zone is obtained by mapping the stirring torque to the apparent viscosity using a mapping function.

[0038] The controller encapsulates material status information into tasks and stores them in the corresponding task queue for processing. For example, it encapsulates real-time solid content and real-time fluid viscosity of materials in each zone and places them in the task queue. In the 120m³ calcium nitrate tetrahydrate crystallization system, the controller extracts attribute tags of task objects from each task queue every cycle. These tags include task object identifiers, which can distinguish between tasks and functional zones. The system also includes material tags such as real-time solid content and real-time fluid viscosity characteristic tags.

[0039] The real-time fluid viscosity mapping logic includes collecting the effective power output of the stirring unit at different frequencies, combining it with the material density obtained from the pressure sensing point in the current partition, and using the correlation model between the fluid flow field Reynolds number and the power criterion number to calculate the dynamic viscosity of the task object in reverse.

[0040] In this embodiment, the diameter D of the bottom stirring blade installed in the 120 cubic meter crystallizer is fixed at 1.2 meters. Based on the physical configuration of the blade, the correlation model constants for this blade type in the transition zone, i.e., when the stirring Reynolds number is between 10 and 104, were obtained through offline clean water experiments: model constant A = 1200, model constant b = 0.5. The critical power number in the fully turbulent zone is 5.0. During a certain scan cycle of continuous crystallization production, the controller extracts the following real-time physical data from multiple sensors and frequency converters on site: Effective power P, such as the frequency converter reading that the real-time output power of the drive motor is 18 kW, after deducting the mechanical transmission no-load loss, the effective power acting on the fluid is P = 14571 W. Stirring speed N, such as the frequency converter reporting that the current stirring speed is 45 revolutions per minute, which is converted to N = 0.75 revolutions per second. Material density ρ, such as the local apparent density ρ = 1450 kg / m³ calculated by the bottom double flange differential pressure transmitter after dynamic pressure compensation.

[0041] The above inverse calculation steps and processes include: Step 1: Determining the flow field state: First, substitute the collected data into the standard definition formula for the power level Np:

[0042] At this point, the calculated power level is approximately 9.574, which is significantly higher than the critical power level of 5.0 for the fully turbulent region. Based on this, the controller determines that the flow field at the bottom of the 120 cubic meter crystallizer is currently in the transitional flow region, where dynamic viscosity still has a significant impact on power consumption, and the model has the conditions for inverse solution.

[0043] Step 2: Reverse substitution into the association model Based on the above association model: Substituting the Reynolds number formula Re = (ρ·N·D²) / η into the equation, we obtain the ultimate derivation formula for the inverse dynamic viscosity η:

[0044] Step 3: Solve for dynamic viscosity Substituting the known data and intermediate results from the first step into the final formula, since b = 0.5, then 1 / b = 2: η ≈ 0.0997 Pa·s. Therefore, the calculated real-time dynamic viscosity of this batch of calcium nitrate tetrahydrate is approximately 0.0997 Pa·s. Seconds. Because the viscosity is set between 80 and 100 millipascals. Within a stable range of seconds, the logic scheduling unit determines that the current fluid flow is good and there is no tendency for local crystallization and agglomeration. Therefore, it continues to maintain the basic quota load output of the stirring unit and does not trigger the second stage of resource gain.

[0045] It should be noted that in this embodiment, the power consumption monitoring point utilizes the power monitoring module built into the inverter of the drive motor included in the stirring system as a virtual sensor to provide real-time feedback on operating current fluctuations. The pressure sensing point employs a corrosion-resistant flange-type diaphragm pressure transmitter, installed on the sidewalls of the core mixing zone and the bottom enrichment zone of the guide tube. Compared to laser particle size analyzers costing hundreds of thousands of yuan, this embodiment achieves millisecond-level monitoring of the crystallization process within a 120 m³ volume using only a pressure gauge costing tens of thousands of yuan and data communication with the inverter. This measurement method provides a 3-5 minute advance warning for fine crystal bursts, leaving ample computational redundancy for subsequent low-energy scheduling of 0.3-0.4 tons.

[0046] In addition, such as Figure 2As shown, the crystallizer includes: a crystallizer body 1 (mixer), a top cover 2, a drive motor 4, a frequency converter 9, a stirring shaft 5, stirring blades 6, a feed inlet, a discharge outlet 13, and an external insulation layer 3. The crystallizer body 1 is made of stainless steel. The drive motor 4, stirring shaft 5, and stirring blades 6 are connected in sequence and installed above the top cover of the crystallizer 2. Solenoid valves are installed at the feed inlet and discharge outlet 13. The feed inlet includes a first feed inlet 11 and a second feed inlet 12. In this embodiment, the stirrer composed of the stirring shaft 5 and stirring blades 6 inside the crystallizer is a multi-layer axial flow stirrer. The stirring blades 6 have at least two layers. The multi-layer axial flow stirrer enhances material dispersion and achieves effective control of crystal particle size. In this embodiment, the stirring speed is controlled by a frequency converter, which is not shown in the figure. Specifically, based on the crystal growth size, the speed of the stirrer is controlled by a frequency converter to minimize power consumption during the crystallization process while ensuring that the crystals do not settle. In other words, the minimum speed at which crystal settling does not occur.

[0047] In addition, a first electrically controlled valve is installed on the feed pipe, and a second electrically controlled valve is installed on the circulation pipe. The controller 10 is electrically connected to the first and second electrically controlled valves, respectively. During operation, the controller 10 controls the first and second electrically controlled valves according to the supersaturation of the mixture in the mixer to adjust the ratio of the product solution to the mother liquor. The water-soluble fertilizer continuous crystallization system provided in this embodiment accurately controls the real-time solid content of the material, as well as the real-time fluid viscosity, temperature, flow rate, and heat exchange rate, through online measurement and automated control technology.

[0048] A first flow meter is installed on the feed pipe, a second flow meter is installed on the circulation pipe, and a third flow meter is installed at the outlet 13 of the crystallizer. The flow meters monitor the feed and discharge flow rates of the crystallizer to ensure that the total feed and discharge rates are equal. In operation, the controller 10 monitors the flow rate of the crystallization stock solution in the feed pipe via the first flow meter. The controller 10 monitors the flow rate of the mother liquor in the circulation pipe via the second flow meter.

[0049] A temperature probe 15, such as a first temperature probe, is connected to the mixer and is electrically connected to the controller 10. The first temperature probe monitors the temperature inside the mixer and sends the monitored temperature information to the controller 10, enabling precise control of the supersaturation of the mixture within the mixer. In use, the controller 10 monitors the temperature of the mixture inside the mixer via the first temperature probe on the mixer.

[0050] Step 3: Drive the logic scheduling unit to determine the basic resource quota required to maintain the steady-state field of material distribution inside the crystallizer based on the hierarchical order of each functional partition; instruct the physical execution unit to perform load output according to the basic resource quota.

[0051] The steady-state field of material distribution is defined by the basic characteristic attributes of each task object.

[0052] The controller drives the logic scheduling unit to determine the scheduling parameters of each functional partition according to the mapping relationship described above, thereby determining the corresponding basic resource quota. It then issues instructions to the physical execution unit according to the mapping relationship. These instructions carry the corresponding basic resource quota and execute the load output corresponding to the mapping relationship based on the resource quota.

[0053] It should be understood that the steady-state field of material distribution is composed of or determined by the material state information of each functional zone. The controller-driven logic scheduling unit obtains task objects from different zones in the task queue and determines the basic resource quota based on the information carried in the task objects. After determining the basic resource quota, the corresponding resource quota is allocated according to the scheduling parameters and mapping relationship as described above.

[0054] Step 4: Real-time analysis of the load attributes of task objects in each queue of tasks to be processed, calculation of the local crystallization load of each functional partition, and determination of the schedulable resource reserve for the current time window based on the difference between the total rated power of the physical execution components and the allocated quota.

[0055] The local crystallization load is obtained by extracting the real-time solid content and real-time fluid viscosity from the load attributes of each task object, and calculating the energy demand deviation caused by the solute precipitation rate in each functional zone using a preset phase change resistance model.

[0056] The aforementioned phase transformation resistance model includes establishing a heat accumulation factor within a functional zone due to the heat exchange rate lagging behind the crystal growth rate, and using the heat accumulation factor as a weighted correction parameter for the local crystallization load.

[0057] For example, the solid content and fluid viscosity mentioned above are input into the built-in phase change resistance model. At a high yield of 5 tons / hour, if the solid content increases and the viscosity also increases, it means that crystal growth has consumed too much solute and local flow resistance has increased. The model calculates the kinetic energy required to maintain particle suspension under the current state and compares it with the basic resource quota allocated in the first stage, producing an energy demand deviation value. This deviation value directly reflects the intensity of the crystallization reaction in the physical world. To further improve the accuracy of perception, a heat accumulation factor is also introduced. The rate of change of the temperature difference between the inlet and outlet of the cooling jacket in this zone and the real-time temperature of the material is monitored. If the rate of decrease in material temperature lags behind the theoretical cooling curve, it indicates that there is latent heat accumulation in the local area, which will induce explosive nucleation. The heat accumulation factor is used as a weighting coefficient to weight and correct the energy demand deviation value, and finally synthesize the local crystallization load, which is determined by the following formula: Local crystallization loading = (phase transformation energy deviation + flow resistance correction) × (1 + heat accumulation factor) Among them, the phase change energy deviation, also known as the main load term, is a core indicator reflecting whether materials are accumulating, and is determined by the following formula: Phase transition energy deviation = (current real-time solid content) (Steady-state standard solid content) / Steady-state standard solid content When the concentration of calcium nitrate tetrahydrate crystals at the bottom of the crystallizer exceeds the designed steady-state value, for example, from 15% to 25%, the deviation increases rapidly, indicating that the material tasks in the task queue have physically accumulated.

[0058] The aforementioned flow resistance correction, also known as the auxiliary load term, is an early warning indicator reflecting whether agglomeration will occur, and is determined by the following formula: Flow resistance correction = (current real-time fluid viscosity) (Standard operating viscosity) / Standard operating viscosity When the viscosity exceeds 150 mPa·s, it indicates that the slurry has thickened or begun to agglomerate, and the stirring unit needs to consume more resources to maintain shear, so the load is increased.

[0059] The aforementioned heat accumulation factor, also known as the gain coefficient, is a multiplier effect reflecting the risk of system runaway, and is determined by the following formula: Heat buildup factor = proportionality coefficient × (actual temperature change rate) (Theoretical rate of change of temperature) / (Theoretical rate of change of temperature) If the actual cooling is too rapid, fine crystal formation may occur; if the cooling is too slow, insufficient heat transfer may occur. In this case, this factor will increase from 0, multiplying the load of the first two factors. This will force the control system to allocate backup resources from the 37kW power source for emergency intervention. This ensures that when heat and solids accumulate simultaneously, the load will increase non-linearly and exponentially, forcibly triggering a higher level of resource scheduling.

[0060] The logical scheduling unit reads the task objects in the queue and parses the load attributes, i.e., the material status information, within the object. It then determines the local crystallization load corresponding to each partition. Based on the material execution component, such as the aforementioned stirring system containing a stirrer, or the total rated power of the stirrer and the allocated power corresponding to the already allocated quota, it determines the power corresponding to the schedulable resource surplus. The stirring system may also include a drive motor.

[0061] Step 5: Monitor the dequeue readiness status of task objects within the underlying functional partition in real time; when the dequeue readiness status meets the preset output confidence threshold, dynamically adjust the opening of the discharge mechanism based on the target capacity so that the corresponding material can be dequeued in a controlled manner.

[0062] The determination of the dequeue readiness state includes real-time calculation of the local apparent density of materials of task objects within the underlying functional partition; when the apparent density is within the preset finished product density range, the dequeue readiness state is determined to have reached the output confidence threshold.

[0063] The opening adjustment strategy of the discharge mechanism includes establishing a time correlation function between the target capacity and the discharge readiness index, and using pulse width modulation logic to dynamically adjust the opening window of the discharge mechanism so that the total amount of material discharged per unit time matches the preset target capacity.

[0064] The opening adjustment of the discharge mechanism adopts a predictive control algorithm, which includes predicting the amount of disturbance to the steady-state field of the underlying functional area during the discharge process based on the correlation curve between historical discharge flow rate and local apparent density. At the same time as the discharge mechanism is opened, the material conveying power from the core mixing zone to the bottom layer is increased in a compensatory manner to offset the negative pressure fluctuation caused by the discharge.

[0065] Local apparent density refers to the mass per unit volume of a mixed fluid composed of the mother liquor (liquid phase), crystals (solid phase), and any possible bubbles (gas phase) within a specific spatial region inside the crystallizer. "Local" emphasizes spatial heterogeneity. In a 120m³ tank, the bottom has more crystals and a higher density, while the top has more clear liquid and a lower density. Apparent density refers to the density distinct from that of a single component. It varies with the solid content of the crystals, such as the slurry concentration. In calcium nitrate tetrahydrate crystallization, as crystals grow, their settling rate increases, and the solid content in the bottom region significantly increases. The increase in local apparent density directly reflects the degree of accumulation and maturity of the bottom crystals.

[0066] In addition, the controlled de-out also includes a feedback adjustment step, including monitoring the steady-state field fluctuation of the material distribution in the bottom functional area during the de-out process. If the local crystallization load of the area is abnormally increased due to the de-out, the logic scheduling unit will trigger emergency resource scheduling to maintain production continuity by increasing the material replenishment rate from the upper functional area to the bottom.

[0067] The dequeue determination and emission control also includes synchronization with the downstream centrifugal separation system, including obtaining real-time processing frequency and filter cake moisture content feedback of the downstream centrifuge through the logic scheduling unit; when the centrifuge is in standby cycle and the dequeue ready state meets the threshold, the discharge mechanism is triggered to perform pulse-type start; if the centrifuge reports an overload signal, the logic scheduling unit forces the task object to be suspended in the queue of pending tasks by lowering the output confidence threshold of the dequeue ready state.

[0068] The method in this embodiment includes continuous feeding of mother liquor and hot raw material liquid, and continuous discharge of crystal slurry. After the crystal slurry is continuously discharged from the bottom outlet of the crystallizer, it directly enters the centrifuge for solid-liquid separation. The feed flow rate is consistent with the crystal slurry discharge flow rate to avoid material accumulation or interruption. The separation process is continuous throughout and synchronized with the crystal slurry discharge rate. The centrifuge is equipped with overload protection and material level monitoring functions. If excessive feed or separation abnormality occurs, the feed flow rate is automatically adjusted or an alarm is triggered to ensure that the separation process is uninterrupted and does not affect the overall continuous crystallization process.

[0069] The centrifuge described in this invention is a horizontal screw discharge sedimentation centrifuge, which is suitable for solid-liquid separation of water-soluble fertilizer slurry. It is corrosion-resistant and scale-resistant, can handle slurry with high solid content, and can discharge continuously without stopping the machine. It solves the drawbacks of traditional centrifuges such as intermittent discharge and small processing capacity, and ensures the smooth progress of continuous production.

[0070] After the centrifuge starts, the internal drum rotates at high speed, ranging from 3000 to 4500 r / min, with a centrifugal factor ≥3000, generating a strong centrifugal force. When the slurry enters the centrifuge drum, under the action of centrifugal force, the denser crystal particles are rapidly thrown against the inner wall of the drum, forming a dense solid layer. The centrifuge's built-in screw discharger rotates in the same direction as the drum but at a different speed, continuously pushing the solid layer from the inner wall of the drum to the centrifuge's solid discharge port, where it is discharged and enters subsequent drying and packaging processes. The less dense mother liquor remains in the liquid phase region inside the drum and is continuously discharged through the centrifuge's liquid outlet, completing efficient solid-liquid separation.

[0071] Furthermore, preferably, the present invention also includes, if the schedulable resource surplus is greater than a preset threshold, the logical scheduling unit dynamically calculates the intra-layer weight based on the local crystallization load of each queue of tasks to be processed, and adds the schedulable resource surplus to the corresponding functional partition according to the intra-layer weight ratio, so as to eliminate the task backlog in the corresponding partition.

[0072] The dynamic resource allocation weight Wi of region i can be defined by the following formula:

[0073] Among them, Q base，i For the basic quota of region i, C real，i C represents the true crystal concentration in region i. target，i Let η be the target crystal concentration in region i. viscosity α is the viscosity influence factor. α and β are known coefficients.

[0074] The aforementioned additional allocation step includes converting the additional resource quota into the instantaneous gain power of the stirring unit or the medium flow rate increment of the cooling unit in the physical execution component. The instantaneous gain power is limited by the total rated power of the physical execution component.

[0075] For example, the controller calculates in real time that the available schedulable resource margin is a total rated power of 37kW – allocated power of 22kW = 15kW. This margin is greater than the system's preset threshold, such as 5kW, and enters the gain-ready state. At this time, through real-time acquisition of material status parameters, it is found that in the task queue of the core mixing zone, due to the instantaneous increase in the amount of raw material liquid replenished, the local solid content surges from 15% to 25%, causing the local crystallization load of this zone to exceed the warning line. The controller immediately starts the weight recalculation program through the logic scheduling unit: the controller calculates the weight within the layer based on the length of the task queue of each zone, such as the number of backlogged task objects, and the slope of change. For example, the weight is 0.7 for the core mixing zone, 0.2 for the annular circulation zone, and 0.1 for the bottom enrichment zone. The 15kW margin is allocated proportionally, and the core mixing zone receives an additional quota of 15kW * 0.7 = 10.5kW. After obtaining the additional quota, it is converted into physical layer control commands through the logic scheduling unit and sent to the corresponding physical execution components. The variable frequency mixer in the core mixing zone received a power compensation command, and its operating power instantly increased from the basic 22kW to 22kW + 10.5kW = 32.5kW.

[0076] At this point, the total output power is 32.5kW, which does not exceed the system's total rated power of 37kW, ensuring the safe operation of the motor. If the load in this area exhibits heat accumulation, the quota will be converted into an increase in the medium flow rate of the cooling unit. By increasing the frequency of the chilled water circulation pump, the heat exchange efficiency will be improved to eliminate the risk of supersaturation.

[0077] The elimination determination of the local crystallization load includes real-time monitoring of the changes in material status information of each functional zone after additional allocation. When the local crystallization load returns to the steady-state field range, the logic scheduling unit recovers the instantaneous gain power and returns this part of the resources to the schedulable resource reserve pool.

[0078] Preferably, the present invention also includes a load reduction monitoring mode. The sensor array continuously feeds back the material status information of the partition. As the rotation speed increases, the shear force strengthens, and the cooling intensifies, the solid phase distribution of the partition gradually becomes more uniform, and the real-time viscosity decreases. When the calculated local crystallization load of the partition returns to the steady-state field range, i.e., the preset normal load range, for three consecutive scan cycles, it is determined that the task accumulation has been eliminated. The additional 10.5kW gain power is withdrawn through the instruction execution component of the logic scheduling unit. This 10.5kW can be re-marked as Idle_Resource and returned to the schedulable resource reserve pool, waiting for the scheduling request of the next high-load partition.

[0079] The present invention preferably also includes a second scheduling phase, namely, adaptive reallocation of remaining resources. If remaining resource quotas are detected within the current time window, for example, if a certain temperature segment meets the standard and no further cooling water supply is needed, the second scheduling phase is triggered. At this time, the intra-layer weight of each layered region, such as the task queue, is calculated. This weight is proportional to the real-time crystal concentration monitored by the online ultrasonic probe and the real-time viscosity measured by the viscosity sensor, such as the current accumulation amount in the queue. If a large amount of crystals precipitates in the lower segment, such as a large accumulation amount, all remaining cooling resources from the first phase or the surplus stirring power margin in the 37kW is immediately allocated to the lower segment to accelerate the final formation of large-particle crystals.

[0080] When applied to the production of calcium nitrate tetrahydrate, this method configures the hierarchical functional zones with specific temperature gradients. The cooling rate gradient from the core mixing zone to the bottom enrichment zone is set to 0.5°C to 1.2°C per hour, triggering the allocation of the aforementioned schedulable resource margin to the cooling unit for gain distribution. A viscosity threshold is set for the calcium nitrate tetrahydrate task object. When the real-time fluid viscosity of the task object exceeds 150 mPa·s, the logic scheduling unit determines that a local crystallization and agglomeration trend has occurred, forcibly increasing the local shear force output of the stirring unit until the viscosity returns to the steady-state range of 80 to 100 mPa·s.

[0081] The system inertia of the 120m³ system, which controls the feed and discharge windows, is significant due to capacity fluctuations. When the actual discharge flow rate of calcium nitrate tetrahydrate, for example, the target 5t / h, deviates from the target value of the crystal slurry discharge flow rate calculated by the controller: if the solid content is too low, for example, below 20%, the system determines that the current batch is immature, directly triggers the window switch parameter indication to close, suspends the second stage of resource allocation to this queue, and simultaneously reduces the opening of the bottom discharge regulating valve, allowing the material to continue queuing and growing in the 120m³ cavity. When the confidence level of the crystallization endpoint determined by the fuzzy control rule library reaches the threshold, the window switch opens, triggering full-power discharge. The fuzzy control rule library comprehensively considers temperature gradient, turbidity, and stirring power.

[0082] By establishing a fuzzy inference rule base, the controller converts the real-time values ​​of the multi-dimensional process parameters into fuzzy variables based on a preset temperature function. It then performs inference calculations on these fuzzy variables according to the fuzzy inference rule base, defining a fuzzy set and its temperature function for each parameter. Substituting each precise value into the corresponding temperature function converts the real-time values ​​into corresponding fuzzy vectors, yielding the fuzzified result. The centroid method is used to calculate the confidence level C of the crystallization endpoint at the current moment. conf When the confidence value of the crystallization endpoint is C conf When the preset first threshold is reached, the controller generates a crystallization endpoint warning signal and records the current time t0.

[0083] For example, the temperature gradient state can be divided into three fuzzy sets: stable temperature gradient (small temperature change), normal temperature gradient (moderate temperature change), and abnormal temperature gradient (large temperature change). When the rate of change of the temperature gradient ΔT = 0.3℃ / h, the temperature function for a stable temperature gradient is defined as follows: When the temperature gradient change rate ΔT = 0.3℃ / h, substituting into the function yields μ = 1.0. Therefore, the temperature gradient change rate ΔT = 0.3℃ / h is converted into vector 1.0, which is completely stable. All process parameters are then sequentially converted into their corresponding fuzzy vectors.

[0084] The fuzzy inference rule base contains multiple fuzzy conditional statements based on expert experience. The preconditions of these fuzzy conditional statements involve fuzzy state combinations of multiple process parameters, and the conclusion is the confidence level of the crystallization endpoint. For example, when the upper and lower segments are temperature gradients T... up -T mid If the turbidity fluctuation is <0.5℃ / h, the turbidity rise rate dTurb / dt <5% / h, and the stirring power dP / dt <2% / h, then the confidence level of the crystallization endpoint is high. When the turbidity fluctuation dTurb / dt >10% / h or the temperature gradient T... up -T midIf the temperature is >1℃, the confidence level for the crystallization endpoint is low. When the sedimentation index S > sedimentation threshold, it indicates that crystals have begun to settle, and the confidence level for the crystallization endpoint is extremely low, requiring sedimentation to be addressed first. If the flow balance deviation ΔQ > 5%, the flow rate should be adjusted first, and endpoint determination should be paused.

[0085] The temperature function is a mathematical function that maps a precise input value to a value between 0 and 1, representing the degree to which the input value belongs to a certain fuzzy set. The temperature function includes trigonometric functions, trapezoidal functions, and Gaussian functions.

[0086] The centroid method is a commonly used method for defuzzification. It calculates the precise value corresponding to the centroid position of the comprehensive fuzzy output set, as shown in the following formula. Where x is the possible value of the output crystallization endpoint confidence vector, ranging from 0 to 100%; It is the membership degree when the confidence value is x after rule-based reasoning.

[0087] In this invention, the first threshold is set to 70%, and the specific value can be customized according to the type of crystal being crystallized.

[0088] Upon generating a pre-alarm, the controller sends a trigger command to the microscopic imaging unit. The automatic sampler extracts crystal slurry from the bottom outlet pipe and sends it into the imaging cavity. The camera captures images of the crystal slurry in the imaging cavity, continuously acquiring three images. The image processing module preprocesses the acquired crystal images, including grayscale conversion and Gaussian filtering for noise reduction. The Canny algorithm is used for edge detection to identify the crystal outline. The equivalent circle diameter method is used to calculate the particle size of each crystal, and the particle size distribution characteristic parameters are statistically obtained: median particle size D50, cumulative 90% particle size D90, and the particle size distribution width D90 / D50 is calculated. The minimum bounding rectangle method is used to calculate the average aspect ratio of the crystal to determine whether it is needle-like or plate-like.

[0089] Meanwhile, to ensure statistical reliability, data was collected three times consecutively and the average value was taken.

[0090] Based on product requirements, set the target D50. target Maximum allowable distribution width D90 / D50 max Maximum allowable aspect ratio AR max The controller compares the particle size distribution characteristic parameters with preset target product specification parameters. When the particle size distribution characteristic parameters simultaneously satisfy D50 ≥ 0.95 × D50... target D90 / D50≤D90 / D50 max Aspect ratio ≤ AR maxWhen the crystallization endpoint is confirmed, the controller sends a command to the bottom discharge regulating valve to start discharging at the flow rate corresponding to the target capacity. If any condition is not met, even if the first-level pre-alarm has been triggered, discharging will be temporarily suspended, and feedback will be sent to the controller to adjust the process parameters: If D50 is too small, it indicates that the crystals are too fine, so the residence time can be appropriately extended by reducing the discharge flow rate or increasing the feed flow rate, or the crystallization temperature can be slightly reduced. If D90 / D50 is too large, it indicates that the particle size distribution is wide, so check for local supersaturation fluctuations and adjust the temperature gradient or stirring speed. If the aspect ratio is too large, it indicates that the crystals are needle-like, so the stirring speed can be appropriately increased.

[0091] Using the particle size distribution characteristics extracted from crystal image analysis as the final judgment basis, it avoids unqualified crystals caused by premature or delayed discharge, ensures that the quality of discharged crystals meets the target specifications in real time, effectively controls the relative standard deviation of median particle size between batches, and stabilizes the particle size distribution width; at the same time, it reduces the frequency of manual sampling and testing, and realizes intelligent and autonomous operation of the crystallization process.

[0092] The following is a detailed description with reference to specific embodiments: Example 1: A vertical cylindrical crystallizer with a height-to-diameter ratio of 3:1 was used to continuously crystallize potassium nitrate water-soluble fertilizer.

[0093] The mother liquor after centrifugation to separate potassium nitrate crystals is continuously added to the crystallizer through the first feed inlet, and the liquid level inside the crystallizer is always kept 30cm above the upper end of the stirring blade.

[0094] The multi-layer axial flow agitator was activated, and the initial viscosity of the material was monitored by a viscosity sensor to be 55 mPa. The initial low speed was adjusted to 38 r / min, and the crystal distribution was monitored in real time by an ultrasonic probe. After the crystal showed a sedimentation trend, the speed was gradually increased to 75 r / min by the frequency converter. The crystal sedimentation trend disappeared. Then, the speed was finely adjusted downward in increments of 2 r / min. When slight crystal sedimentation was detected at 73 r / min, 75 r / min was locked as the minimum effective speed. During the production process, dual-parameter linkage control was used to maintain the speed stability.

[0095] A hot feedstock solution composed of water, potassium nitrate, and magnesium nitrate is continuously added to the crystallizer through a second inlet. The mother liquor feed flow rate is monitored at 3 m³ / s using an electromagnetic flowmeter. 3 / h, hot raw material liquid feed flow rate is 12m 3 / h, total feed flow rate is 15m³ / h 3The controller adjusts the opening of the discharge regulating valve to maintain a dynamic balance between the crystal slurry discharge flow rate and the total feed flow rate. Simultaneously, the internal heat exchanger is divided into three independent heat exchange zones: upper:middle:lower = 2:3:5. The target constant temperature T for the material inside the crystallizer is set to 25℃. The cooling medium temperature is controlled at 22℃ for the upper zone, 20℃ for the middle zone, and 18℃ for the lower zone. The cooling medium flow rate in each zone is adjusted in real time according to the crystal concentration, with the lower zone having the highest crystal concentration and a cooling medium flow rate of 1.0 m³ / h. 3 / (h m 2 The controller precisely adjusts the temperature and flow rate of the cooling medium to maintain a constant temperature of the material inside the crystallizer, thus enabling isothermal crystallization.

[0096] The target production capacity is set at 600 kg / h, which translates to a target crystal slurry discharge flow rate of 15 m³ / h. 3 / h, the controller monitors the actual crystal slurry discharge flow rate in real time, compares it with the target value and dynamically adjusts the opening of the discharge regulating valve, and simultaneously adjusts the hot raw material liquid feed flow rate to maintain a stable crystal slurry discharge flow rate; the crystal slurry is sent to a horizontal screw discharge sedimentation centrifuge for solid-liquid separation, the crystal particles are pushed out by the screw discharger, and the separated mother liquor is returned to the first feed port of the crystallizer for recycling.

[0097] In this embodiment, the particle size uniformity of the potassium nitrate water-soluble fertilizer crystal product reaches over 92%, the system energy consumption is reduced by 9% compared to the traditional continuous crystallization method, the heat exchanger heat exchange efficiency is increased by 12%, and low-energy, continuous large-scale production is achieved, with no significant quality difference between product batches.

[0098] Example 2: A vertical cylindrical crystallizer with a height-to-diameter ratio of 2:1 was used to continuously crystallize potassium dihydrogen phosphate water-soluble fertilizer.

[0099] The mother liquor after centrifugally separating potassium dihydrogen phosphate crystals is continuously added to the crystallizer through the first feed inlet, and the liquid level inside the crystallizer is always kept 20cm above the upper end of the stirring blade.

[0100] The multi-layer axial flow agitator was activated, and the initial viscosity of the material was monitored by a viscosity sensor to be 35 mPa. The initial low speed was adjusted to 25 r / min, and the crystal distribution was monitored in real time by an ultrasonic probe. After a slight sedimentation trend was detected, the speed was gradually increased to 80 r / min by a frequency converter. The sedimentation trend disappeared. Then, the speed was finely adjusted downward in increments of 5 r / min. When a slight sedimentation of the crystal was detected at 75 r / min, 80 r / min was locked as the minimum effective speed. During the production process, dual-parameter linkage control was used to maintain the speed stability.

[0101] A hot feedstock solution composed of water and potassium dihydrogen phosphate is continuously added to the crystallizer through the second inlet. The mother liquor feed flow rate is monitored by an electromagnetic flowmeter and is 2 m³ / s. 3 / h, hot raw material liquid feed flow rate is 8m 3 / h, total feed flow rate is 10m³ / h 3 The controller adjusts the opening of the discharge regulating valve to maintain a dynamic balance between the crystal slurry discharge flow rate and the total feed flow rate. Simultaneously, the internal heat exchanger is divided into three independent heat exchange zones: upper, middle, and lower sections in a ratio of 2:3:5. The target constant temperature T for the material inside the crystallizer is set at 35℃. The cooling medium temperature is controlled at 32℃ in the upper section, 30℃ in the middle section, and 28℃ in the lower section. The cooling medium flow rate in each section is adjusted in real time according to the crystal concentration, with the lower section having the highest crystal concentration and a flow rate adjusted to 1.2 m³ / h. 3 / (h m 2 The controller precisely adjusts the temperature and flow rate of the cooling medium to maintain a constant temperature of the material inside the crystallizer, thus enabling isothermal crystallization.

[0102] The target production capacity is set at 500 kg / h, which translates to a target crystal slurry discharge flow rate of 10 m³ / h. 3 / h, the controller monitors the actual crystal slurry discharge flow rate in real time, compares it with the target value and dynamically adjusts the opening of the discharge regulating valve, and simultaneously adjusts the hot raw material liquid feed flow rate to maintain a stable crystal slurry discharge flow rate; the crystal slurry is sent to a horizontal screw discharge sedimentation centrifuge for solid-liquid separation, the crystal particles are pushed out by the screw discharger, and the separated mother liquor is returned to the first feed port of the crystallizer for recycling.

[0103] In this embodiment, the particle size uniformity of potassium dihydrogen phosphate crystals reaches over 90%, the system energy consumption is reduced by 8% compared to the traditional continuous crystallization method, the heat exchanger heat exchange efficiency is increased by 10%, and there are no problems such as crystal sedimentation or large-scale precipitation of fine crystals during the production process, resulting in stable product quality.

[0104] Taking the crystallization production of potassium dihydrogen phosphate water-soluble fertilizer as the subject, a control group was set up using the existing conventional continuous crystallization method, and the experimental group used the crystallization method of the present invention. Both groups used crystallization equipment of the same specifications, the same raw material liquid concentration and target production capacity, as shown in Table 1. The production energy consumption, crystal product quality and production efficiency of the two groups were compared.

[0105] Table 1 Control group and experimental group

[0106] The results of the comparative experiments show that the crystallization method of the present invention is significantly superior to the existing conventional continuous crystallization method in terms of production energy consumption, crystal product quality, and production efficiency. It can effectively solve many drawbacks of the existing technology and has significant technical advantages and industrial application value.

[0107] This method also includes firstly, dynamically adjusting the stirring blades to the minimum effective speed, and then controlling the speed in conjunction with the dual parameters of material viscosity and crystal concentration. This minimizes the energy consumption of stirring while completely avoiding crystal sedimentation. Simultaneously, the multi-layered axial-flow stirring blades drive material circulation, ensuring uniform mixing within the crystallizer and providing a stable environment for crystal growth. Secondly, the internal heat exchanger employs a segmented temperature gradient control structure, scientifically controlling the temperature in different zones based on the crystal concentration distribution and stirring flow field characteristics within the crystallizer. This effectively suppresses the precipitation of fine crystals in the upper section and promotes uniform crystal growth in the lower section, significantly improving crystal particle size uniformity and stabilizing product quality. Furthermore, the flow rate in each section is precisely adjusted according to the crystal concentration, avoiding ineffective consumption of cooling medium, improving heat exchange efficiency, and further reducing system energy consumption. Thirdly, an electromagnetic flowmeter achieves dynamic balance between the total feed and crystal slurry discharge flow rates. The feed and discharge flow rates are adjusted in conjunction with the target production capacity to achieve precise capacity control, adapting to different production plan requirements. The separated mother liquor is recycled, reducing raw material waste and further lowering overall production costs. The crystallizer used in the end has a simple structure, and all supporting equipment are mature industrial equipment. It is easy to operate and maintain, and does not require complicated manual operation. It realizes the automated control of the crystallization process, reduces the intensity of manual operation, and has low manufacturing and operating costs, and has good prospects for industrial application.

[0108] This invention enables continuous crystallization production of water-soluble fertilizers, eliminating the batch operation of traditional intermittent crystallization and removing time losses in auxiliary feeding and discharging processes. It significantly improves equipment utilization and production efficiency, with a unit volume capacity far exceeding existing crystallization equipment, making it suitable for large-scale industrial production. The traditional simple water or steam control of crystallizers is upgraded to the allocation of resource groups of different computing power types, transforming the 37kW motor power and the 0.3-0.4 tons of steam per ton of product into a dynamic quota library. Overcoming control lag in large systems, a two-stage scheduling method is directly applied to a 120m³ mega-system. The first stage maintains the basic flow and temperature fields; the second stage uses crystal concentration and viscosity as the current accumulation amount in the request queue to dynamically extract surplus steam heat energy or idle motor power, precisely targeting the areas with the most severe crystal accumulation. This not only completely solves the problem of localized bottoming but also greatly reduces steam consumption to the extreme range of 0.3-0.4 tons.

[0109] It should also be noted that the crystallizer mentioned above includes a main support unit, a stirring drive unit, a heat exchange and temperature control unit, a material inlet and outlet unit, and an online monitoring unit.

[0110] The aforementioned main support unit is the basic framework of the crystallizer, used to provide material storage space and support all functional components, such as... Figure 2As shown, the system includes the aforementioned body 1, top cover 2, and external insulation layer 3. The body 1 is a vertical cylindrical, hollow stainless steel cavity, vertically fixed to the production base, forming a space for material crystallization. The top cover 2 is a circular stainless steel end cap structure adapted to the body 1, sealed to the top opening of the body 1 via flange bolts, achieving a sealed environment for the crystallizer and preventing material evaporation and the entry of external impurities. The external insulation layer 3 covers the outer wall of the body 1 and the outer surface of the top cover 2, fitting tightly against them. The inner side of the external insulation layer 3 is made of heat-insulating material, while the outer side is a protective shell, used to reduce heat exchange between the crystallizer and the outside environment, helping to maintain a constant internal temperature. The mother liquor after centrifugal crystal separation is continuously added to the crystallizer through the first feed inlet via the main support unit, ensuring that the liquid level inside the crystallizer is always higher than the upper surface of the stirring blades 6. The aforementioned stirring drive unit is used to drive the material circulation and prevent crystal sedimentation. It includes a drive motor 4 (motor), a frequency converter, a stirring shaft 5, and multi-layer axial flow stirring blades. The stirring shaft 5 and the multi-layer axial flow stirring blades can form a stirrer, arranged along the vertical central axis of the crystallizer. The frequency converter is not shown in the figure. Stirring is started by the stirring drive unit, and the stirring blade speed is dynamically adjusted to the minimum effective speed while ensuring that the crystals do not settle. The drive motor 4 is coaxially and fixedly connected to the frequency converter, and is installed as a whole at the top center of the top cover 2 to provide driving force for stirring. The input end of the frequency converter is rigidly connected to the output shaft of the drive motor 4, and the output end passes through the top cover 2 and extends into the interior of the crystallizer, driving the stirring shaft 5 to rotate. The stirring shaft 5 is a long stainless steel shaft. The upper end is sealed to the output end of the frequency converter through a coupling, and the lower end extends vertically to the lower area of ​​the crystallizer body 1, arranged coaxially with the body 1. The stirring blades 6 have a multi-layered axial flow structure, preferably 2 to 4 layers, and are fixedly sleeved at equal intervals along the axial direction of the stirring shaft 5, with the entire structure located within the effective liquid height range of the crystallizer body 1. The uppermost stirring blade 6 has its upper end face lower than the normal operating liquid level of the crystallizer, and the lower end face of the lowest stirring blade 6 maintains a distance of 5 to 10 cm from the bottom of the body 1, ensuring that the stirring blades 6 are completely immersed in the material and do not contact the body 1.

[0111] The heat exchange and temperature control unit and the material inlet and outlet unit enable the hot raw material liquid to be crystallized to be continuously added to the crystallizer through the second inlet 12. The total feed and crystal slurry discharge flow rates are monitored in real time and kept in dynamic balance. At the same time, the temperature and flow rate of the cooling medium in the internal heat exchanger are monitored in real time and automatically adjusted to keep the material temperature in the crystallizer constant and carry out constant temperature crystallization.

[0112] The heat exchange and temperature control unit is used to precisely control the crystallization temperature. It includes an internal heat exchanger, a circulating pump 8, and an external heat exchanger 7, adopting an internal heat exchange and external circulation structure. The crystallizer also includes a frequency converter 9, which, along with the internal heat exchanger, circulating pump 8, and external heat exchanger 7, realizes an internal and external circulating cooling loop. The internal heat exchanger is a stainless steel tubular heat exchanger, in the shape of an annular coil, divided into three independent heat exchange areas: upper, middle, and lower. It is coaxially sleeved around the stirring blades 6, and the minimum distance between it and the inner wall of the crystallizer body 1 is not less than 5 cm. In this invention, the internal heat exchanger is divided into three independent heat exchange areas along the effective liquid height of the crystallizer in a ratio of upper:middle:lower = 2:3:5. Each area is equipped with an independent temperature probe 15, flow sensor, and regulating valve. The inlet and outlet of the internal heat exchanger penetrate the outer wall of the body 1 and extend to the outside of the crystallizer. The internal heat exchanger is used to introduce a cooling medium to remove the heat brought in by the raw material liquid and the heat released by crystallization. The outlet of the internal heat exchanger is connected to the inlet of the circulating pump 8 via a pipeline. The outlet of the circulating pump 8 is connected to the inlet of the external heat exchanger 7, and the outlet of the external heat exchanger 7 is connected back to the inlet of the internal heat exchanger via a pipeline, forming a closed cooling medium circulation loop. Both the circulating pump 8 and the external heat exchanger 7 are electrically connected to the controller 10. A regulating valve is installed on the cooling medium circulation pipeline, integrated with the heat exchange temperature control unit, to achieve precise control of the cooling medium temperature and flow rate.

[0113] The material inlet / outlet unit enables continuous feeding of mother liquor and hot raw material liquid, and continuous discharge of crystal slurry. It includes a first inlet 11, a second inlet 12, and a bottom outlet 13, each independently arranged. The second inlet 12 is a circular flange interface, located on one side of the top cover 2 of the crystallizer. Its outer end is connected to the outlet of the mother liquor pump 17 via a pipeline, and its inner end extends into the crystallizer and faces the stirring shaft 5, used to continuously feed the circulating mother liquor in the mother liquor tank 18 into the crystallizer. The first inlet 11 is also a circular flange interface, located on the other side of the top cover 2 of the crystallizer, symmetrically arranged with the second inlet 12. Its outer end is connected to the hot raw material liquid storage tank via a pipeline, and its inner end extends into the crystallizer, used to continuously feed the hot raw material liquid to be crystallized into the crystallizer. The bottom outlet 13 is a circular flange interface, located at the center of the bottom of the crystallizer body 1. Its inner end communicates with the interior of the body 1, and its outer end is connected to the inlet of the centrifuge via a pipeline, used to continuously discharge the crystal slurry from the crystallizer. Electromagnetic flow meters are connected in series on the conveying pipelines of the first feed inlet 11, the second feed inlet 12, and the bottom discharge outlet 13. The electromagnetic flow meters are electrically connected to the controller to realize real-time monitoring and dynamic balance control of the inlet and outlet flow rates.

[0114] The independent arrangement of the first feed inlet 11 and the second feed inlet 12 effectively prevents the mother liquor and hot raw material liquid from mixing prematurely in the pipeline, ensuring that the two materials are fully mixed by the stirring blades 6 in the crystallizer before crystallization, thus preventing local oversaturation or uneven crystallization. Furthermore, the independent design of the first feed inlet 11 and the second feed inlet 12 facilitates cleaning and maintenance of the mother liquor feed pipeline and the hot raw material liquid feed pipeline without shutting down the equipment, improving equipment maintainability.

[0115] The mother liquor feed pipeline and the hot raw material feed pipeline are each equipped with an electromagnetic flow meter, which can independently adjust the mother liquor feed flow rate and the hot raw material feed flow rate, providing a flexible means of operation for liquid level control and production capacity adjustment.

[0116] The online monitoring unit is used to collect process parameters within the crystallizer in real time, providing data support for automatic control. It includes a level gauge 14, a temperature probe 15, and an ultrasonic probe 16, each positioned at a different location on the crystallizer body. The level gauge 14 is a differential pressure or radar-type level gauge. Its probe end is installed at the front end of the crystallizer top cover 2, extending into the crystallizer interior. Its signal output is connected to the controller 10 via a signal line, used to monitor the material level within the crystallizer in real time, ensuring the level is always higher than the upper surface of the stirring blade 6. The temperature probe 15 is a stainless steel armored temperature sensor, installed in the middle of the outer wall of the crystallizer body 1. Its probe end penetrates the sidewall and is inserted into the material. Its signal output is connected to the controller 10, used to monitor the material temperature in real time. The ultrasonic probe 16 is the core component for crystal concentration detection, including an upper probe and a lower probe. Both probe ends penetrate the sidewall and are inserted into the material, facing the stirring shaft. Its signal output is connected to the controller 10, used to monitor the crystal content and distribution in the material in real time. The upper probe is installed on the upper part of the outer wall of the crystallizer body, corresponding to the upper end face of the stirring blade, and the lower probe is installed on the lower part of the outer wall of the crystallizer body, corresponding to the lower end face of the stirring blade.

[0117] The drive motor 4 is electrically connected to the frequency converter 9. The signal terminals of the frequency converter 9, electromagnetic flow meter, liquid level gauge 14, temperature probe 15, ultrasonic probe 16, circulating pump 8, and external heat exchanger 7 are all connected to the controller 10 to realize the full-process automated control of crystallizer stirring speed, feed and discharge flow rate, heat exchange temperature, and crystal concentration.

[0118] More specifically, the mother liquor from the centrifuge solid-liquid separation is first temporarily stored in the mother liquor tank 18, then pressurized and transported by the mother liquor pump 17, and continuously injected into the internal cavity of the crystallizer through the pipeline into the second feed port 12 of the crystallizer. The mother liquor feed is a continuous flow, forming a closed-loop material circulation with the subsequent crystal slurry discharge and hot raw material liquid feed, ensuring uninterrupted operation of the crystallization process.

[0119] The mother liquor is a partially crystallized water-soluble fertilizer salt solution that still contains crystallizable active ingredients. Recycling it avoids raw material waste, maximizes raw material utilization, and reduces the raw material cost per unit product. Furthermore, the composition and temperature of the mother liquor are highly consistent with the materials in the crystallizer. Compared to directly adding fresh, cold raw materials, this avoids drastic disturbances to the crystallization environment caused by new materials, reduces energy consumption for subsequent cooling / heating, and ensures a stable crystal growth environment.

[0120] The level gauge 14, installed on the top cover 2 of the crystallizer, continuously detects the actual liquid level of the material in the crystallizer and transmits the data to the controller.

[0121] The controller 10 compares the real-time liquid level in the crystallizer with the height of the upper surface of the stirring blade 6. If the real-time liquid level is lower than the upper surface of the stirring blade 6, the stirring blade 6 will be partially or completely exposed to the gas phase, resulting in idling, a sharp drop in stirring efficiency, insufficient material circulation, and easy crystal settling and agglomeration. This also causes load fluctuations in the drive motor 4, generating additional energy consumption and mechanical wear. Therefore, when the level gauge 14 malfunctions or the liquid level exceeds the upper or lower limits, an alarm is triggered and the feed is automatically cut off to prevent the crystallizer from drying out or overflowing, ensuring production safety.

[0122] Specifically, when the real-time liquid level is lower than the upper surface of the stirring blade 6, it is determined to be too low. The controller sends a liquid level adjustment command to the electromagnetic flow meter or regulating valve on the mother liquor feed pipe to increase the mother liquor feed flow rate and quickly raise the liquid level in the crystallizer. When the real-time liquid level is higher than the upper surface of the stirring blade 6, it is determined to be normal, and the current feed flow rate is maintained to keep the dynamic stability. When the real-time liquid level is close to the upper limit of the crystallizer, it is determined to be too high. The controller sends a liquid level adjustment command to the electromagnetic flow meter or regulating valve on the mother liquor feed pipe to reduce the mother liquor feed flow rate and prevent liquid overflow.

[0123] In actual production, the liquid level needs to be maintained at 5-30 cm above the upper surface of the stirring blade 6, ensuring complete submersion of the blade while allowing sufficient gas phase space to prevent material overflow. The mother liquor feed flow rate varies depending on the crystallizer specifications and production capacity, and is typically controlled between 0.5 and 5 m³ / h. 3 / h, specifically adjusted automatically by the controller based on the real-time liquid level.

[0124] More specifically, the drive motor 4, installed directly above the top cover 2 of the crystallizer, is activated. The power from the drive motor 4 is reduced in speed by a frequency converter and then transmitted to the stirring shaft 5, driving the multi-layered axial-flow stirring blades 6, which are fixedly connected to the stirring shaft 5, to rotate. The stirring shaft 5 is arranged along the vertical central axis of the crystallizer, and the stirring blades 6 are located within the effective liquid height range of the crystallizer and are completely submerged below the stable liquid level. This causes the material to circulate up and down within the crystallizer, ensuring that the crystals are uniformly suspended in the material and preventing crystal deposition and agglomeration. Simultaneously, it enhances the heat and mass transfer processes between materials, ensuring uniform temperature and concentration distribution within the crystallizer, providing a stable and uniform environment for uniform crystal growth.

[0125] An ultrasonic probe 16 is used to monitor crystal settling in real time, preventing crystals from settling to the bottom of the crystallizer due to gravity. This avoids crystal accumulation at the bottom of the crystallizer, which can lead to agglomeration, adhesion to the inner wall of the equipment, or even blockage of the bottom outlet 13, ensuring continuous production. Simultaneously, it ensures that all crystals are in a suitable flow field and temperature environment for growth, effectively improving crystal particle size uniformity and product quality stability.

[0126] The ultrasonic probe 16 collects real-time crystal distribution data of the material in the crystallizer. When a sudden increase in local crystal concentration, excessive concentration difference between the upper and lower probes, or abnormal fluctuations in the ultrasonic signal are detected, if the difference between the crystal concentration detected by the lower probe and the crystal concentration detected by the upper probe exceeds a preset concentration threshold for more than 30 seconds, or if the crystal concentration detected by the lower probe changes within 10 seconds, it is determined that crystal sedimentation has begun. Otherwise, if the crystal concentration remains stable, the concentrations of the upper and lower layers are equal, or the ultrasonic signal remains unchanged, it is determined that crystal sedimentation has not occurred.

[0127] To determine whether crystal sedimentation has occurred, manual verification is used. Specifically, samples are taken from the top, middle, and bottom of the crystallizer every 1-2 hours. If crystals accumulate at the bottom and the content in the upper layer is significantly lower, then sedimentation has been confirmed. Otherwise, no sedimentation has occurred.

[0128] However, the method of adjusting the stirring speed by only the crystal concentration does not take into account the characteristic that the viscosity of water-soluble fertilizer materials changes with the amount of crystal precipitation. This can easily lead to local crystal sedimentation caused by the lag in speed adjustment, or energy waste caused by excessive speed.

[0129] Therefore, this invention, based on single-parameter speed regulation of crystal concentration, introduces online monitoring of material viscosity and achieves dual-parameter linkage. A variable frequency drive (VFD) is electrically connected to the drive motor 4, and the VFD performs closed-loop speed regulation based on the monitoring data. Specifically, the stirring blade 6 speed is first adjusted to the initial low-speed threshold range of 20~40 r / min. Simultaneously, a viscosity sensor is introduced to monitor the real-time viscosity of the material. When the material viscosity ≥ 50 mPa... When the viscosity is <50 mPa, the initial low rotation speed is increased to 35-40 r / min to counteract the flow resistance caused by high viscosity. At time s, maintain the initial low speed of 20~30 r / min to reduce energy consumption.

[0130] When the ultrasonic probe 16 detects a sedimentation trend in the crystals, the frequency converter gradually increases the rotational speed in increments of 2-5 r / min until the sedimentation trend completely disappears. This speed is recorded as the critical speed. Based on the critical speed, the speed is then finely adjusted downwards in increments of 2-5 r / min. When a slight sedimentation trend is detected, the speed of the previous increment is locked as the minimum effective speed. The overall range of the minimum effective speed is 50-120 r / min, with 80-120 r / min for potassium sulfate and potassium dihydrogen phosphate fertilizers, and 50-80 r / min for potassium nitrate and calcium ammonium nitrate fertilizers. When the material viscosity is ≥50 mPa... At time s, adjust the gradient to 2 r / min, when the material viscosity is <50 mPa. At time s, adjust the gradient to 5 r / min.

[0131] During continuous crystallization production, if the ultrasonic probe detects a change in the crystal distribution, the variable frequency speed controller will dynamically adjust the speed in real time to always maintain the stirring blade 6 at the locked minimum effective speed, achieving the dual goals of energy saving and stable suspension.

[0132] By linking and controlling both viscosity and crystal concentration, precise and proactive adjustment of stirring speed is achieved, avoiding crystal sedimentation caused by changes in material viscosity. This further reduces stirring power consumption, balancing crystallization stability with low energy consumption. An adjustment gradient is also set, allowing for flexible adaptation to different crystallization needs based on the characteristics of water-soluble fertilizer crystals. This solves the problem of poor speed adaptability during crystallization of different fertilizers, improving versatility.

[0133] More specifically, the hot raw material liquid to be crystallized is a high-temperature saturated / supersaturated mixture, prepared by mixing water and water-soluble fertilizer salts according to the process ratio. It is temporarily stored in a temperature-controlled hot raw material liquid storage tank, maintaining the temperature of the raw material liquid 10-20°C higher than the target constant temperature of the crystallizer, providing a basis for cooling crystallization. The hot raw material liquid is pressurized and transported by the hot raw material liquid storage tank discharge pump, and continuously fed into the first feed port 11 of the crystallizer through the hot raw material liquid pipeline. The first feed port 11 is an independent flange interface opened on the top cover 2 of the crystallizer, symmetrically arranged with the second feed port 12, with its inner end extending into the interior of the crystallizer cavity and facing the direction of the stirring shaft 5, forming a counter-feed with the mother liquor. After the hot raw material liquid enters the crystallizer in a continuous flow state, it is rapidly and uniformly mixed with the continuously added mother liquor under the high-speed circulation stirring of the multi-layer axial flow stirring blades 6, avoiding sudden changes in local material concentration and temperature, and providing a uniform material basis for the crystallization reaction.

[0134] The aforementioned water-soluble fertilizer salts include one or more of the following commercially available salt components: potassium sulfate, magnesium sulfate, ammonium sulfate, potassium dihydrogen phosphate, ammonium dihydrogen phosphate, potassium nitrate, calcium ammonium nitrate, calcium nitrate, magnesium nitrate, potassium chloride, and magnesium chloride.

[0135] The total feed is the sum of the flow rates of the mother liquor feed at the first feed port 11 and the hot raw material feed at the second feed port 12. The crystal slurry discharge flow rate is the crystal slurry outflow rate at the bottom outlet of the crystallizer. The dynamic balance between the two is the core guarantee for continuous crystallization.

[0136] Specifically, electromagnetic flowmeters are connected in series on the conveying pipelines of the first feed inlet 11 and the second feed inlet 12, as well as on the crystal slurry discharge pipeline at the bottom outlet of the crystallizer. These three electromagnetic flowmeters collect real-time data on the mother liquor feed flow rate, the hot raw material feed flow rate, and the crystal slurry discharge flow rate, and transmit all flow data synchronously to the controller 10. The controller 10 automatically calculates the total feed flow rate and compares it with the crystal slurry discharge flow rate in real time to determine the flow difference. Based on the flow difference, the controller sends control commands to the feed solenoid valve and the bottom discharge regulating valve of the mother liquor feed pipeline or the hot raw material feed pipeline. When the total feed flow rate > the crystal slurry discharge flow rate, the opening of the bottom discharge regulating valve is increased to increase the crystal slurry discharge flow rate, while the opening of the feed solenoid valve can be finely adjusted to decrease the hot raw material or mother liquor feed flow rate. When the total feed flow rate < the crystal slurry discharge flow rate, the opening of the bottom discharge regulating valve is decreased to decrease the crystal slurry discharge flow rate, while the opening of the feed solenoid valve is finely adjusted to increase based on the liquid level. When the total feed flow rate equals the crystal slurry discharge flow rate, all valve openings are kept constant to maintain dynamic flow stability. The flow balance control is linked with the controller 10. When the flow adjustment causes a small fluctuation in the liquid level in the crystallizer, the controller 10 quickly corrects the liquid level by fine-tuning the mother liquor feed flow rate to ensure that the real-time liquid level in the crystallizer is always higher than the upper end face of the stirring blade 6.

[0137] The cooling medium of the internal heat exchanger is cooling circulating water. The internal heat exchanger is used to exchange heat and cool down the material in the crystallizer. The precise adjustment of its temperature and flow rate is accomplished by the heat exchange temperature control unit.

[0138] Temperature probe 15, located in the middle of the side wall of the crystallizer body 1, collects the actual temperature of the material inside the crystallizer in real time, serving as the core basis for temperature control. Temperature probe 15 and flow sensor are installed on the inlet and outlet pipes of the cooling medium in the internal heat exchanger, respectively, to collect the inlet and outlet temperatures and circulation flow rate of the cooling medium in real time. All temperature and flow data are transmitted to controller 10 in real time. Controller 10 compares the actual temperature of the material with the preset target constant temperature value T of the material inside the crystallizer, and simultaneously combines this with the temperature and flow rate data of the cooling medium to determine whether the heat exchange efficiency matches the crystallization requirements, and synchronously executes temperature and flow rate adjustments for the cooling medium.

[0139] The existing integrated temperature control method of internal heat exchanger is prone to uneven heat exchange in different height areas of the crystallizer. The lower material has a higher temperature due to high crystal concentration and insufficient heat exchange, resulting in a slow crystallization rate. The upper material has excessive heat exchange, resulting in a large amount of fine crystal precipitation, ultimately causing uneven crystal particle size distribution and unstable product quality.

[0140] To address technical issues encountered during use, this invention controls the operating parameters of the cooling medium refrigeration or heating device, adjusting the initial temperature of the cooling medium to ensure that the temperature of the cooling medium entering the internal heat exchanger matches the crystallization requirements. By adjusting the opening of the cooling medium inlet regulating valve on the inlet pipe of the internal heat exchanger, the circulation flow rate of the cooling medium is changed, achieving precise control of the heat exchange power.

[0141] The controller 10 independently controls the temperature of the upper, middle, and lower sections. The cooling medium temperature is controlled at T-3 to T-5℃ for the upper section, T-5 to T-7℃ for the middle section, and T-7 to T-10℃ for the lower section. The cooling medium flow rate for each section is adjusted in real time according to the crystal concentration of the corresponding area; the higher the crystal concentration, the greater the cooling medium flow rate. T represents the average temperature of the material in the crystallizer. In this invention, the cooling medium value ranges from 0.8 to 1.5 m³. 3 / (h m 2 ,Hour (square meters)

[0142] By segmented gradient temperature control and precise adjustment of the cooling medium, combined with the external insulation layer 3 on the outside of the crystallizer body 1 and the enhanced internal heat transfer of the material through stirring and homogenization, the crystal concentration distribution characteristics of the upper, middle and lower sections of the crystallizer are adapted to achieve precise gradient heat exchange, effectively suppressing the precipitation of fine crystals in the upper part and promoting the uniform growth of crystals in the lower part, thereby improving the uniformity of crystal particle size and increasing the product qualification rate; at the same time, the segmented flow rate is precisely adjusted to avoid ineffective consumption of cooling medium and improve the overall heat exchange efficiency of the heat exchanger.

[0143] More specifically, according to the production plan, the target production capacity is preset in the controller 10, that is, the mass of water-soluble fertilizer crystal products to be produced per unit time; combined with the preset crystal slurry solid content, that is, the mass ratio of crystals in the crystal slurry, the crystal slurry solid content is generally controlled at 20%~35% depending on the fertilizer type, as shown in the following formula, the target production capacity is converted into the corresponding crystal slurry discharge flow target value: Crystal slurry discharge flow target value = target production capacity ÷ crystal slurry solid content ÷ crystal slurry density.

[0144] The target value of the crystal slurry discharge flow rate is synchronously transmitted by the controller 10 to the flow sensor and discharge regulating valve at the bottom outlet, and simultaneously links the total feed and crystal slurry discharge flow rates to dynamically balance, forming a dual closed-loop control, which serves as the core benchmark for flow regulation.

[0145] Specifically, a flow sensor installed on the crystal slurry discharge pipeline at the bottom outlet of the crystallizer collects the actual crystal slurry discharge flow rate data in real time, updates the data every second, and transmits it synchronously to the controller 10. The controller 10 compares the actual crystal slurry discharge flow rate with the preset crystal slurry discharge flow rate target value in real time and calculates the flow difference. At the same time, based on the flow difference and the total feed flow rate data, it sends an electrical signal to the discharge regulating valve at the bottom outlet to automatically adjust the valve opening.

[0146] When the actual crystal slurry discharge flow rate is less than the preset target value, increase the opening of the discharge regulating valve to increase the discharge flow rate. The adjustment range is positively correlated with the flow rate difference, with each adjustment ranging from 5% to 10% until the actual flow rate matches the target value. Simultaneously, increase the hot raw material liquid feed flow rate at the second inlet 12 to ensure a balance between the total feed and discharge flow rates. When the actual crystal slurry discharge flow rate is greater than the preset target value, decrease the opening of the discharge regulating valve to reduce the discharge flow rate. Simultaneously, reduce the hot raw material liquid feed flow rate to prevent insufficient total feed from causing a drop in liquid level. When the actual crystal slurry discharge flow rate equals the preset target value, maintain the current discharge regulating valve opening and hot raw material liquid feed flow rate to ensure stability.

[0147] After adjustment, the controller 10 continuously monitors the deviation between the actual flow rate and the target value. When the deviation is ≤ ±2%, the opening of the discharge regulating valve remains unchanged. When the deviation exceeds the range, the above adjustment process is repeated to ensure that the crystal slurry discharge flow rate is always accurately matched with the target capacity.

[0148] The separated mother liquor is transported to a mother liquor collection tank through a liquid phase outlet pipeline for temporary storage and simple filtration to remove any small amounts of entrained fine crystals, preventing impurities from entering the crystallizer and affecting product quality. The mother liquor in the collection tank is pressurized by mother liquor pump 17 and transported to the second feed inlet 12, where it merges with new centrifugally separated mother liquor and is continuously added to the crystallizer to participate in the crystallization reaction again. The mother liquor circulation flow rate is linked to the mother liquor feed flow rate and jointly controlled by controller 10 to ensure that the amount of circulating mother liquor added does not affect the stability of the liquid level and material balance within the crystallizer.

[0149] However, controlling the output flow rate of crystal slurry solely by converting the target capacity without considering the dynamic fluctuations in the solid content of the crystal slurry during the crystallization process leads to a significant deviation between the actual crystal output and the target capacity, resulting in a lag in capacity control. Furthermore, fluctuations in solid content cause unstable centrifuge separation loads, easily leading to incomplete separation, crystal entrainment with mother liquor, or mother liquor entrainment with fine crystals, thus reducing product purity and raw material utilization.

[0150] Therefore, a new online real-time monitoring and dynamic linkage control mechanism for the solid content of the crystal slurry is added. An online solid content detector can be connected in series on the crystal slurry discharge pipe at the discharge port 13 at the bottom of the crystallizer. This online solid content detector is electrically connected to the controller 10 and collects the actual solid content data of the crystal slurry in real time, updating the data every 0.5 seconds. The controller 10 simultaneously substitutes the target capacity and real-time solid content into the conversion formula to dynamically correct the target value of the crystal slurry discharge flow rate: Corrected target value of crystal slurry discharge flow rate = Target capacity ÷ Real-time solid content ÷ Crystal slurry density.

[0151] The online solids content detector's probe is tightly fitted to the inner wall of the crystal slurry discharge pipeline, and the probe area avoids bends and valve positions in the crystal slurry discharge pipeline to ensure the accuracy of the detection data. The controller 10 has a built-in solids content abnormality alarm module. When the real-time crystal slurry solids content exceeds the range of 20%~35%, an audible and visual alarm is triggered, and the cooling medium flow rate and the opening of the discharge regulating valve are automatically adjusted. If the adjustment still fails to restore it to the preset range, the controller 10 automatically cuts off the feed and shuts down for protection. The online solids content detector of this invention uses the microwave transmission method, which is suitable for the characteristics of water-soluble fertilizer crystal slurry.

[0152] The controller 10 simultaneously compares the actual crystal slurry discharge flow rate with the corrected target value, and the real-time crystal slurry solid content with the preset crystal slurry solid content range, performing dual adjustments. Specifically, when the real-time crystal slurry solid content > 35%, it indicates that there are too many crystals in the crystallizer. The opening of the discharge regulating valve is increased, and the cooling medium flow rate is increased simultaneously to accelerate the crystallization rate and prevent crystal accumulation. When the real-time crystal slurry solid content < 20%, it indicates that there are too few crystals in the crystallizer. The opening of the discharge regulating valve is decreased, and the cooling medium flow rate is decreased simultaneously to prolong the crystal growth time. When 20% < real-time crystal slurry solid content < 35%, the current discharge regulating valve opening and cooling medium flow rate are maintained.

[0153] Simultaneously, the online solid content data is transmitted to the controller 10. The centrifuge automatically adjusts the drum speed according to the real-time solid content. For every 5% increase in the actual solid content of the crystal slurry, the drum speed is increased by 500 r / min to ensure that the solid-liquid separation efficiency matches the state of the crystal slurry.

[0154] By adding a new online real-time monitoring and dynamic linkage control mechanism for the solid content of crystal slurry, the accuracy of capacity control has been significantly improved, completely solving the problem of lagging capacity control and adapting to the precise capacity requirements of industrial-scale production; secondly, the centrifuge speed is dynamically adapted to the solid content of crystal slurry to avoid the problem of incomplete separation, thereby improving the purity of crystal products and further enhancing the utilization rate of raw materials.

[0155] However, when using isothermal crystallization, the timing of crystal growth completion and discharge often depends on empirical timing or a single parameter, which can easily lead to a high misjudgment rate and the risk of missed judgments due to process fluctuations.

[0156] A turbidimeter, a stirring motor power monitoring module, and a microscopic imaging unit are added to the crystallizer to collect multi-dimensional process parameters reflecting the state of the crystallization process in real time. These multi-dimensional process parameters include the upper section temperature T. up Mid-section temperature T mid Lower section temperature T low Turbidity value of crystal slurry (Turb), stirring power (P), and feed / discharge flow rate balance deviation (ΔQ=Q) in -Q ou t, ultrasonic probe settlement index S.

[0157] The turbidimeter is installed on the middle side wall of the crystallizer, with its probe immersed in the material to collect turbidity data of the crystal slurry in real time. This turbidity data reflects the crystal content and particle size variation trend in the slurry, and the signal is connected to controller 10. The stirring motor power monitoring module is integrated into the frequency converter to collect stirring motor power data in real time, indirectly reflecting changes in slurry viscosity. The microscopic imaging unit is installed on a branch road beside the bottom discharge port and includes an automatic sampler, an imaging chamber, a camera, and an image processing module. The inlet of the automatic sampler is connected to the bottom discharge port pipeline, and the outlet is connected to the imaging chamber. The imaging chamber has a transparent flow channel structure. The camera is located outside the imaging chamber and electrically connected to the image processing module. The signal output terminal of the image processing module is connected to the controller. The camera is a high-resolution CCD industrial camera with more than 5 million pixels, equipped with a microscope lens, used to capture crystal images. The image processing module has built-in edge detection and particle size analysis algorithms, outputting characteristic values ​​such as the average crystal particle size and particle size distribution width in real time.

[0158] The above description illustrates preferred embodiments of the present invention and helps those skilled in the art to more fully understand the technical solution of the present invention. However, these embodiments are merely illustrative and should not be construed as limiting the specific implementation of the present invention to these embodiments. For those skilled in the art, several simple deductions and modifications can be made without departing from the inventive concept, and all such modifications should be considered within the protection scope of the present invention.

Claims

1. A dynamically scheduled method for crystallizing water-soluble fertilizers, applied in a continuous crystallization apparatus, characterized in that, The method includes the following steps: Establish a mapping relationship between logical scheduling units and physical execution components in the controller, and configure scheduling parameters that characterize the weight of resource quota distribution; The internal space of the crystallizer is divided into multiple functional zones with a hierarchical order, and the material status information of each functional zone is encapsulated as a task object and stored in the corresponding task queue to be processed. The logic scheduling unit is driven to determine the basic resource quota required to maintain a steady-state material distribution field inside the crystallizer based on the hierarchical order of each functional partition; the steady-state material distribution field is defined by the basic characteristic attributes of each task object; and the physical execution unit is instructed to perform load output according to the basic resource quota. The load attributes of the task objects in each queue of tasks to be processed are analyzed in real time, the local crystallization load of each functional partition is calculated, and the schedulable resource margin of the current time window is determined based on the difference between the total rated power of the physical execution components and the allocated quota. The local crystallization loading is determined by the following formula: Local crystallization loading = (phase change energy deviation + flow resistance correction) × (1 + heat accumulation factor); The phase transition energy deviation is determined by the following formula: Phase transition energy deviation = (current real-time solid content - steady-state standard solid content) / steady-state standard solid content; The flow resistance correction is determined by the following formula: Flow resistance correction = (current real-time fluid viscosity) (Standard operating viscosity) / Standard operating viscosity; The heat accumulation factor is determined by the following formula: Heat buildup factor = proportionality coefficient × (actual temperature change rate) (Theoretical temperature change rate) / (Theoretical temperature change rate) The system monitors the dequeue readiness status of task objects within the underlying functional partitions in real time. When the dequeue readiness status meets the preset output confidence threshold, the opening of the discharge mechanism is dynamically adjusted based on the target capacity to ensure controlled dequeueing of the corresponding materials.

2. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 1, characterized in that, It also includes a method whereby if the schedulable resource balance is greater than a preset threshold, the logical scheduling unit dynamically calculates the intra-layer weight based on the local crystallization load of each queue of tasks to be processed, and adds the schedulable resource balance to the corresponding functional partition according to the intra-layer weight ratio, so as to eliminate task backlog in the corresponding partition.

3. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 2, characterized in that, It also includes converting additional schedulable resource quotas into instantaneous gain power of the stirring unit or medium flow rate increment of the cooling unit in the physical execution component; wherein the instantaneous gain power is limited by the total rated power of the physical execution component.

4. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 3, characterized in that, It also includes a logic scheduling unit that monitors the changes in material status information of each functional zone after additional allocation in real time. When the local crystallization load returns to the steady-state range, the logic scheduling unit recovers the instantaneous gain power and returns this part of the resources to the schedulable resource reserve pool.

5. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 1, characterized in that, It also includes the logical scheduling unit extracting the real-time solid content and real-time fluid viscosity from the load attributes of each task object, and calculating the energy demand deviation caused by the solute precipitation rate in each functional zone through a preset phase change resistance model.

6. A dynamically scheduled crystallization method for water-soluble fertilizers according to claim 5, characterized in that, The preset phase transformation resistance model includes: establishing a heat accumulation factor within the functional zone due to the heat exchange rate lagging behind the crystal growth rate, and using the heat accumulation factor as a weighted correction parameter for the local crystallization load.

7. A method for dynamically scheduling water-soluble fertilizer crystallization according to claim 6, characterized in that, The method for obtaining the material status information includes: setting pressure sensing points and power consumption monitoring points in each of the functional zones, calculating the real-time solid content using static pressure difference, and mapping the real-time fluid viscosity using the operating current fluctuation of the physical actuator.

8. A dynamically scheduled crystallization method for water-soluble fertilizers according to claim 1, characterized in that, The determination of the dequeue readiness state includes: real-time calculation of the local apparent density of materials of task objects within the underlying functional partition; when the apparent density is within a preset finished product density range, the dequeue readiness state is determined to have reached the output confidence threshold.

9. A method for dynamically scheduling water-soluble fertilizer crystallization according to claim 8, characterized in that, The opening adjustment strategy of the discharge mechanism includes: establishing a time correlation function between the target capacity and the discharge readiness index, and using pulse width modulation logic to dynamically adjust the opening window period of the discharge mechanism so that the total amount of material discharged per unit time matches the preset target capacity.

10. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 9, characterized in that, The controlled discharge of the corresponding materials also includes monitoring the steady-state fluctuations of the material distribution in the bottom functional area during the discharge process. If the discharge causes an abnormal increase in the local crystallization load of the area, the logic scheduling unit will trigger emergency resource scheduling to maintain production continuity by increasing the material replenishment rate from the upper functional area to the lower level.

11. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 1, characterized in that, The functional zones are divided into a core mixing zone, annular circulation zone, and bottom enrichment zone based on the physical location of the internal guide tube of the crystallizer.

12. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 1, characterized in that, The departure determination and emission control also include pacing matching with the downstream centrifugal separation system: The logic scheduling unit obtains feedback on the real-time processing frequency and filter cake moisture content of the downstream centrifuge; when the centrifuge is in a standby cycle and the dequeue ready state meets the threshold, the discharge mechanism is triggered to perform pulse-type start; if the centrifuge reports an overload signal, the logic scheduling unit forces the task object to be suspended in the queue of pending tasks by lowering the output confidence threshold of the dequeue ready state.

13. A dynamically scheduled water-soluble fertilizer crystallization method according to claim 1 or 11, characterized in that, The opening adjustment of the discharge mechanism is based on the correlation curve between historical discharge flow rate and local apparent density to predict the amount of disturbance to the steady-state field of the underlying functional area during the discharge process; while the discharge mechanism is open, the material conveying power from the core mixing zone to the underlying layer is increased in a compensatory manner to offset the negative pressure fluctuations caused by the discharge.

14. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 7, characterized in that, The mapping logic for the real-time fluid viscosity includes: The effective power output of the stirring unit at different frequencies is collected, and combined with the material density obtained from the pressure sensing point in the current partition, the dynamic viscosity of the task object is obtained by reverse calculation through the correlation model of fluid flow field Reynolds number and power criterion number.

15. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 1, characterized in that, When applied to the production of calcium nitrate tetrahydrate, the hierarchical functional partitions are configured with specific temperature gradients: The cooling rate gradient from the core mixing region to the bottom enrichment region is set to be 0.5 degrees Celsius per hour to 1.2 degrees Celsius per hour; When the local crystallization load shows a cooling rate deviation exceeding ±10%, the available schedulable resources are triggered to perform gain allocation to the cooling unit.

16. The method for dynamically scheduling water-soluble fertilizer crystallization according to claim 15, characterized in that, Set the viscosity threshold for the calcium nitrate tetrahydrate task object: When the real-time fluid viscosity of the target object exceeds 150 mPa·s, the logic scheduling unit determines that a local crystallization and agglomeration trend has occurred, and forcibly increases the local shear force output of the stirring unit until the viscosity returns to the steady-state range of 80 to 100 mPa·s.

Images