Intelligent gradient-based temperature control method for whole-process preparation of urea-based compound fertilizer

Abstract

The application provides a whole-process preparation method of urine-based compound fertilizer based on intelligent gradient temperature control, comprising the following steps: S1, collecting real-time temperature data and heat flow distribution data of each process area in the workshop, and generating a temperature gradient distribution map; S2, performing partition processing on the heat transfer path according to the temperature gradient distribution map, and determining the spatial boundaries of the high-temperature zone, the medium-temperature zone and the low-temperature zone; S3, if the spatial boundaries show that the heat transfer appears disordered characteristics, then the temperature unstable area is classified and processed to obtain the position coordinates of the potential heat interference source; S4, inputting the position coordinates of the potential heat interference source into a three-dimensional layout optimization model to obtain an adjusted three-dimensional workshop structure diagram; S5, extracting interval isolation parameters, and determining the setting parameters of the control valve through heat exchange simulation; S6, performing production process simulation, judging whether the particle cracking risk is lower than the cracking threshold value, and if yes, then the current three-dimensional workshop structure diagram and the corresponding control valve setting parameters are determined as the optimized temperature control scheme.

Description

Technical Field

[0001] This invention relates to the field of intelligent compound fertilizer production technology, and in particular to a complete process preparation method for urea-based compound fertilizer based on intelligent gradient temperature control. Background Technology

[0002] In the fields of agricultural production and fertilizer manufacturing, the production process of urea-based compound fertilizers plays a crucial role, directly impacting crop yields and the sustainable use of soil. However, despite the extremely high demands on product quality and production efficiency in this field, existing technologies often face numerous challenges in practical operation, affecting overall production results.

[0003] Currently, many production methods struggle to adapt to the dynamic changes in heat distribution in complex workshop environments, resulting in inaccurate temperature management. Especially during production, heat transfer and distribution are affected by various factors, and traditional control methods often fail to identify the root cause of problems in a timely manner or effectively respond to sudden temperature fluctuations. This limitation threatens the quality stability of the production process, thereby affecting the performance of the final product.

[0004] A deeper problem lies in the fact that the disorder and complexity of heat transfer have become the core factors affecting temperature control. Heat distribution within the workshop is often uneven; some areas may be overheated, while others are underheated. This imbalance not only increases energy consumption but also damages the product. For example, in the granulation stage of urea-based compound fertilizer, if the temperature in one area is too high, micro-cracks may appear inside the granules due to uneven thermal stress, while areas with excessively low temperatures may result in incomplete granule formation, affecting strength and performance. This contradiction caused by uneven heat distribution directly increases the risk of granule cracking during production and makes it difficult to balance energy consumption and quality.

[0005] Therefore, how to achieve orderly heat transfer and precise control in a complex production environment, and reduce the risk of cracking caused by uneven temperature of granules, has become a key issue in improving the production efficiency and product quality of urea-based compound fertilizer. Summary of the Invention

[0006] To address the technical problems mentioned in the background section, this invention provides a method for the entire process preparation of urea-based compound fertilizer based on intelligent gradient temperature control, comprising: S1, collecting real-time temperature data and heat flow distribution data of each process area in the workshop to generate a temperature gradient distribution map; S2, dividing the heat transfer path into zones according to the temperature gradient distribution map to determine the spatial boundaries of high-temperature, medium-temperature, and low-temperature zones; S3, if the spatial boundaries show disordered heat transfer characteristics, classifying the temperature unstable areas to obtain the location coordinates of potential heat interference sources; S4, inputting the location coordinates of potential heat interference sources into a three-dimensional layout optimization model to obtain an adjusted three-dimensional workshop structure diagram; S5, extracting interval isolation parameters and determining the setting parameters of the control valves through heat exchange simulation; S6, performing production process simulation to determine whether the risk of particle cracking is lower than the cracking threshold. If it is lower, the current three-dimensional workshop structure diagram and the corresponding control valve setting parameters are determined as the optimized temperature control scheme.

[0007] Furthermore, step S1 includes:

[0008] The raw data collected by the sensor network is first filtered to remove impulse noise, and then spatially interpolated to form point cloud data of temperature field and heat flow field covering the entire three-dimensional space of the workshop. Then, the temperature change rate and heat flow vector direction at each spatial point are calculated by the gradient operator, and finally a temperature gradient distribution map is generated.

[0009] Furthermore, step S2 includes: step S21, grouping the spatial points of the heat transfer path in the temperature gradient distribution map; step S22, defining the three-dimensional spatial range of the high-temperature zone, the medium-temperature zone, and the low-temperature zone based on the grouping results; and step S23, using the three-dimensional spatial range as the basis for the regional division of subsequent workshop structure adjustments.

[0010] Furthermore, step S21 includes:

[0011] A density-based spatial clustering method is used to group spatial points along heat transfer paths.

[0012] Furthermore, step S3 includes: step S31, extracting the temperature fluctuation feature vector of the temperature unstable region; step S32, inputting the temperature fluctuation feature vector into the support vector machine model; and step S33, outputting the set of location coordinates of potential thermal interference sources through the support vector machine model.

[0013] Furthermore, step S4 includes: step S41, using the location coordinates of potential heat interference sources, the current equipment placement constraints in the workshop, and the heat transfer direction constraints as input parameters for the three-dimensional layout optimization model; step S42, calculating feasible heat transfer paths with the goal of minimizing the ordered heat transfer path and heat interference; and step S43, adjusting the workshop equipment layout based on the feasible heat transfer paths and outputting the adjusted three-dimensional workshop structure diagram.

[0014] Furthermore, step S42 includes:

[0015] A multi-objective weighted optimization method is adopted, which linearly combines the two objectives of minimizing the orderly heat transfer path and minimizing thermal interference through weight coefficients to form a single objective function.

[0016] Furthermore, step S5 includes: step S51, extracting the isolation distance and isolation material parameters of adjacent functional areas from the adjusted three-dimensional workshop structure diagram; step S52, calculating the controllable heat exchange coefficient between each adjacent area based on the isolation distance, the isolation material parameters, and the temperature difference between adjacent areas; and step S53, determining the installation position and opening setting value of the control valve based on the controllable heat exchange coefficient.

[0017] Furthermore, step S53 includes: step S531, establishing a mapping relationship between the controllable heat exchange coefficient and the opening degree of the control valve; step S532, calculating the opening set value of each control valve through the mapping relationship; and step S533, associating the installation position of the control valve with the opening set value to form complete control valve setting parameters.

[0018] Further, step S6 includes: step S61, generating the temperature-time distribution for each process interval according to the setting parameters of the control valve; step S62, calculating the changes in thermal stress inside the particles sequentially for the processes of urine spraying, material mixing, granulation, drying, and cooling; step S63, determining the particle cracking risk value by comparing the changes in thermal stress with the critical cracking stress threshold of the particle material; and step S64, comparing the particle cracking risk value with the cracking threshold and outputting the judgment result.

[0019] The technical solution provided by this invention has the following beneficial effects:

[0020] This invention discloses a complete preparation method for urea-based compound fertilizer based on intelligent gradient temperature control. Addressing the unique challenges of uneven heat distribution, difficulty in locating heat interference sources, and high risk of particle cracking in urea-based compound fertilizer production, it proposes a systematic solution. The disordered and complex nature of heat transfer in the production workshop makes temperature control difficult, thus affecting product quality. This invention uses a sensor network to collect temperature and heat flow data in real time, generating a temperature gradient distribution map to accurately divide high-temperature, medium-temperature, and low-temperature zones. It then utilizes clustering algorithms and support vector machine models to locate potential heat interference sources, and combines a three-dimensional layout optimization model to adjust the workshop equipment layout and optimize heat transfer paths. Simultaneously, through heat exchange simulation and production process simulation, valve parameters are finely adjusted to effectively reduce particle thermal stress and control cracking risk. Ultimately, this invention achieves the technical effects of orderly heat utilization, a significant reduction in particle crack rate, and optimized energy consumption, comprehensively improving production efficiency and product quality. Attached Figure Description

[0021] Figure 1 This is a flowchart of the whole process preparation method of urea-based compound fertilizer based on intelligent gradient temperature control according to the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0023] like Figure 1 As shown, this embodiment of the invention provides a temperature control method for the preparation of urea-based compound fertilizer based on thermal field sensing. The method includes the following steps: S1, collecting real-time temperature data and heat flow distribution data of each process area in the workshop, and generating a temperature gradient distribution map. Specifically, various types of data are collected in real time through a sensor network.

[0024] In one embodiment, the workshop typically houses multiple sets of temperature and heat flux sensors. These sensors are arranged according to the production process layout in the urine spraying area, material mixing area, granulation area, drying area, cooling area, and transition areas between these areas. Real-time temperature data is primarily collected by thermocouples or infrared thermometers, typically every 5 to 30 seconds, with the specific frequency adjusted according to the production line cycle. Heat flux distribution data is obtained using heat flux meters positioned near key heat transfer surfaces, such as the outer wall of the drying drum, the inner wall of the cooling duct, and near the granulator casing, to characterize the intensity and direction of heat transfer per unit area. In actual production environments, urine-based compound fertilizer preparation workshops often exhibit significant non-uniform heat distribution. The urine spraying process introduces a large amount of latent heat release, the drying process continuously consumes a large amount of heat for moisture evaporation, and the cooling process needs to remove residual heat. These processes form a complex heat transfer network through various pathways such as air convection, equipment radiation, and material conduction. The raw data collected by the sensor network is first filtered to remove impulse noise, and then spatially interpolated to form point cloud data of temperature and heat flow fields covering the entire three-dimensional space of the workshop. Based on this, the rate of temperature change and heat flow vector direction at each spatial point are calculated using gradient operators, ultimately generating a temperature gradient distribution map. This map visually displays high-temperature concentration areas, rapid heat flow channels, and temperature abrupt change zones using color depth or vector arrows. For example, on a production line with an annual output of 300,000 tons of urea-based compound fertilizer, the sensor network is deployed with 248 temperature measuring points and 76 heat flow measuring points. The collected data is spatially meshed to form a 0.5m × 0.5m × 0.5m resolution voxel grid, with each voxel storing the instantaneous temperature value and three-dimensional heat flow vector at that location. In a sampling at a certain moment, a significant high-temperature cluster appeared in the area 0.8m to 2.2m above the urine spraying tower, with a maximum temperature of 78.4℃, while the temperature in the adjacent mixing machine area was only 42.1℃, forming a steep temperature gradient zone between the two. This gradient distribution map provides fundamental data support for subsequent heat transfer path analysis. S2, Based on the temperature gradient distribution map, the heat transfer paths are partitioned to determine the spatial boundaries of high-temperature, medium-temperature, and low-temperature zones. The partitioning of heat transfer paths first extracts features from the spatial points in the temperature gradient distribution map, and then uses an appropriate clustering method to divide the regions. The clustering process uses temperature value, temperature gradient magnitude, and heat flow vector direction as the main clustering criteria.

[0025] Optionally, this step also includes: step S21, grouping the spatial points of heat transfer paths in the temperature gradient distribution map. Specifically, a density-based spatial clustering method can be used to group the spatial points of heat transfer paths. This method can effectively identify thermal clusters of arbitrary shapes without forcibly segmenting the data into spherical or convex regions.

[0026] In one possible implementation, temperature thresholds T1 and T2 are first set, where T1 distinguishes between high-temperature and medium-temperature zones, and T2 distinguishes between medium-temperature and low-temperature zones. For example, T1 can be 55℃ and T2 can be 38℃. Then, the consistency of the heat flow vector direction is used as a second weight to determine the connectivity of adjacent voxels. Only when two adjacent voxels simultaneously satisfy the condition of consistent temperature range affiliation and the angle between their heat flow vectors is less than a preset angle threshold, such as 45°, are they considered to belong to the same heat transfer connectivity branch. After completing the initial grouping in this way, the grouping results are then morphologically processed to remove isolated clumps with excessively small areas and fill small holes, ultimately resulting in multiple independent thermal zone clusters. Step S22: The three-dimensional spatial range of the high-temperature, medium-temperature, and low-temperature zones is defined based on the grouping results. After the above grouping process, each cluster is assigned a category label: the high-temperature zone corresponds to a region with a temperature consistently higher than T1 and a large heat flow intensity; the medium-temperature zone corresponds to a region with a temperature between T2 and T1 and a relatively stable heat flow direction; and the low-temperature zone corresponds to a region with a temperature lower than T2. Boundary determination employs the α-shape algorithm or convex hull algorithm to envelop the outer contour of each category cluster, obtaining a polyhedral boundary description in three-dimensional space. To improve boundary accuracy, locations with larger temperature gradient moduli can be introduced as priority boundary points, ensuring that the defined region boundaries closely approximate locations of drastic actual heat changes. Step S23 uses the defined three-dimensional spatial range as the basis for subsequent workshop structure adjustments. The defined high-temperature, medium-temperature, and low-temperature three-dimensional spatial ranges are saved as basic partition templates for subsequent optimization. For example, the high-temperature zone is mainly concentrated near the urine spraying tower, the granulation drum cavity, and the primary drying section; the medium-temperature zone covers the secondary drying section, screening section, and part of the conveyor belt area; and the low-temperature zone is mainly located at the cooler outlet, finished product conveying, and packaging sections. These spatial ranges are used as constraints for locating thermal interference sources, reconfiguring equipment, and designing thermal isolation in subsequent steps. In actual production, factors such as the moisture content of different batches of raw materials, urea content, and ambient temperature and humidity can cause dynamic changes in the thermal field distribution. For example, when the urea content in the raw material increases from 46% to 52%, the heat of reaction released during the urea spraying stage increases significantly, the volume of the high-temperature zone may expand by 18%–25%, and the transition zone between the medium-temperature and high-temperature zones widens. This dynamic change necessitates real-time updates to the zoning results; steps S1 and S2 are typically re-executed every 10–30 minutes to ensure that the zoning criteria always reflect the current true thermal field state. In step S3, if the spatial boundaries show disordered heat transfer characteristics, the temperature-unstable regions are classified to obtain the location coordinates of potential thermal interference sources. Disordered heat transfer is indicated when adjacent zones exhibit severely jagged boundaries, the temperature gradient direction frequently reverses at the boundaries, or multiple temperature islands appear within the same zone. This disorder is typically caused by factors such as abnormal local heat sources, equipment heat leakage, duct blockage, material accumulation, or steam pipe leakage.

[0027] Optionally, this step also includes: Step S31, extracting the temperature fluctuation feature vector of the temperature unstable region. For the boundary regions and isolated regions determined to be disordered, a local three-dimensional cube window containing the region is first defined, for example, the window size can be 3m × 3m × 4m. Then, within this window, temperature and heat flow sequences are continuously collected for 5 to 15 minutes at time intervals of 0.3 seconds to 1 second. Based on these time series, various statistical features are calculated, including but not limited to: the mean, variance, skewness, and kurtosis of the temperature sequence; the maximum, minimum, and average rate of change of temperature; the angular variance of the heat flow vector direction; the value of the temperature autocorrelation function under different time delays; and frequency domain features such as the dominant frequency position and energy proportion of the power spectrum. The above features are combined to form a multi-dimensional vector, namely the temperature fluctuation feature vector. In one embodiment, the vector has 24 dimensions, of which the first 12 dimensions are temperature statistical features, and the last 12 dimensions are heat flow and frequency domain features. Step S32, inputting the temperature fluctuation feature vector into the support vector machine model. The Support Vector Machine (SVM) model is used as a binary classifier to determine whether the current window contains a significant thermal interference source. The model uses a radial basis function as the kernel function, and the penalty parameter C and kernel parameter γ are obtained through cross-validation with historical data. During the training phase, the model has learned a classification hyperplane between normal fluctuations and abnormal interference based on a large number of labeled samples. Step S33 outputs the set of location coordinates of potential thermal interference sources through the SVM model. Specifically, when the classification result is "significant thermal interference source exists," the weighted centers of multiple candidate locations, such as the point of maximum temperature gradient, the point of sudden change in heat flux intensity, and the point of maximum variance in the temperature sequence, are further calculated as representative coordinates of potential thermal interference sources within the window. If multiple independent interference features are detected simultaneously within a window, multiple coordinate points are output to form a set of location coordinates. For example, in a certain production process, the boundary of the transition section between the granulator and the primary dryer exhibits obvious disordered characteristics: the boundary line shows multiple sharp protrusions in three-dimensional space, and the local temperature rapidly rises from 51.2℃ to 67.8℃ and then falls back to 43.5℃ within a short period. After extracting the temperature fluctuation feature vector of the region, it was input into a trained support vector machine, and the model output determined that "a thermal interference source exists." Further calculation yielded the coordinates of three main interference sources, located 0.7 meters below the granulator outlet, near the roller bearing of the transition conveyor belt, and 1.2 meters to the left of the air inlet of the primary dryer. These locations were subsequently confirmed as roller bearing overheating, conveyor belt frictional heat generation, and air inlet steam leakage, respectively. It should be noted that step S3 is triggered only when the spatial boundary exhibits disordered features. If the partition boundary is smooth and the heat flow direction is highly consistent, S3 can be skipped, and the current partition result can be directly used to enter the subsequent layout optimization process. This selective execution mechanism effectively reduces unnecessary computational overhead and improves the system's real-time response capability.In another possible implementation, surface temperature images acquired by an infrared thermal imager can be further combined to verify and correct the coordinate set output by the support vector machine. Infrared images can intuitively reflect abnormal temperature points on the outer surface of equipment. When the distance between the coordinates predicted by the support vector machine and the high-temperature abnormal point in the infrared image is less than 0.8 meters, the coordinates are retained; otherwise, their confidence is reduced or they are discarded. This multi-source data fusion method significantly improves the accuracy of locating potential thermal interference sources, especially suitable for scenarios such as steam pipe leaks and motor overloads that are difficult to locate directly and accurately through the volumetric temperature field. In one embodiment, the set of location coordinates of potential thermal interference sources obtained in the aforementioned steps, along with the existing equipment placement constraints and heat transfer direction constraints in the current workshop, are used together in the next step of layout adjustment calculation. S4, the location coordinates of potential thermal interference sources are input into the three-dimensional layout optimization model to obtain the adjusted three-dimensional workshop structure diagram. The three-dimensional layout optimization model uses the three-dimensional space of the workshop as the computational domain, treats the equipment as a rigid geometric body with attributes such as volume, heat dissipation power, heat capacity, and surface emissivity, and expresses the heat transfer direction constraints as a series of soft or hard directional weights. The location coordinates of potential thermal interference sources are marked as high-priority avoidance points or mandatory isolation points.

[0028] Optionally, this step also includes: Step S41, using the location coordinates of potential heat interference sources, current workshop equipment placement constraints, and heat transfer direction constraints as input parameters for the 3D layout optimization model. Current workshop equipment placement constraints include, but are not limited to: the urine spraying tower must be located in the north-west area of ​​the workshop for urine pipeline access; a minimum process distance of 2.8 meters must be maintained between the granulation drum and the dryer; the cooler discharge end must be close to the finished product silo conveyor belt inlet; and the main steam main and condensate return pipe trenches must have fixed directions. Heat transfer direction constraints primarily aim to achieve a roughly unidirectional heat flow from the urine spraying area → mixing area → granulation area → drying area → cooling area, avoiding heat backflow or cross-disturbance. In actual input, these constraints are transformed into feasible regions in 3D space, polyhedral obstacles, directional penalty functions, etc. For example, heat transfer direction constraints can be achieved by adding a term with a negative cosine similarity to the expected direction of the heat flow vector to the objective function, making the optimization process tend to align the main heat flow path with the process flow direction. Step S42: Calculate feasible heat transfer paths with the objectives of minimizing the ordered heat transfer path and minimizing thermal interference. The objective of minimizing the ordered heat transfer path means minimizing the geometric length of the main heat flow path from the high-temperature source to the low-temperature heat dissipation end in space, while simultaneously reducing the proportion of the path traversing the medium-temperature and low-temperature zones, all while satisfying all hard constraints. Minimizing thermal interference is achieved by minimizing the thermal radiation and convection coupling strength between potential thermal interference sources and key equipment. The coupling strength can be approximated by a negative exponential function of distance or a weighted average of the squared distance. In one possible implementation, a multi-objective weighted optimization method is used, linearly combining the two objectives of minimizing the ordered heat transfer path and minimizing thermal interference through weighting coefficients to form a single objective function. The weighting coefficients can be dynamically adjusted based on seasonal ambient temperature, urea content fluctuations, production load rates, etc. For example, when the workshop ambient temperature is high and the cooling load is large in summer, the weight of the "minimize thermal interference" objective can be appropriately increased; while in winter, when the raw material moisture content is high and the drying heat consumption is large, the weight of the "shortest ordered transfer path" objective can be appropriately increased. Step S43: Adjust the workshop equipment layout based on the feasible heat transfer path and output the adjusted three-dimensional workshop structure diagram. Specifically, the adjustment process typically involves: translating, rotating, locally raising or lowering some non-critical equipment, fine-tuning the conveyor belt tilt angle, relocating the air duct interface position, and adding heat insulation screens. For example, moving an auxiliary vibrating screen from the right side of the original dryer's tail to the left rear position to avoid the heat recirculation zone of the primary dryer's exhaust fan; changing the air inlet louvers of the secondary cooler from side air inlet to top vertical air inlet to reduce mixing with the hot air from the granulation zone; and adding a movable heat insulation curtain between the primary and secondary dryers.After adjustments, a new 3D digital model of the workshop structure is generated. This model includes the new position coordinates and orientation information of all major equipment, pipelines, air ducts, insulation components, and reserved installation points for control valves. This model is typically stored in a standard 3D format for easy retrieval in subsequent heat exchange simulations and production process simulations. For example, in the technical renovation of a 500,000-ton-per-year urea-based compound fertilizer plant, the space above the granulator in the original layout was occupied by a steam header spanning the workshop, resulting in a stable thermal dome in that area, making it difficult to effectively dissipate heat downstream. The 3D layout optimization model recommended rerouting the steam header along the east side of the workshop's exterior wall, while simultaneously raising the existing exhaust fan unit above the granulator by 1.2 meters and shifting it 0.9 meters westward. After these adjustments, the length of the main heat flow path was shortened by approximately 14%, while the local thermal interference intensity in the granulator area decreased by approximately 31%. In another implementation, when workshop space is extremely limited and numerous hard constraints restrict the conventional adjustment space, a modular, movable base design for some equipment can be introduced. For example, some vibrating screening equipment and belt conveyor transfer stations can be installed on track-mounted bases, allowing them to move laterally or longitudinally within a certain range, thus enabling fine-tuning of the thermal field without changing the position of the main large equipment. S5, extract the interval isolation parameters and determine the setting parameters of the control valves through heat exchange simulation. Interval isolation parameters mainly refer to the physical isolation characteristics used to control heat exchange between adjacent functional intervals, including isolation distance, thermal resistance of the isolation material, surface emissivity, and whether forced convection isolation is installed.

[0029] Optionally, this step also includes: Step S51, extracting the isolation distance and isolation material parameters of adjacent functional areas from the adjusted 3D workshop structure diagram. The isolation distance refers to the shortest straight-line distance or the equivalent distance along the main heat flow direction between the geometric boundaries of two adjacent functional areas. The isolation material parameters include the thickness, thermal conductivity, specific heat capacity, surface emissivity, and other physical properties of heat insulation boards, heat insulation curtains, metal reflectors, air curtains, etc. These parameters are partly derived from design drawings and partly from on-site measurements or data manuals provided by suppliers. For example, the original design net distance between the granulation area and the primary drying area was 3.2 meters, with a 50 mm thick aluminum silicate fiber heat insulation board in the middle, and surface emissivity of 0.87 and 0.32 on both sides, respectively; after adjustment, the net distance increased to 3.9 meters, and a new liftable double-layer aluminum foil reflector was added, reducing the surface emissivity to 0.12. Step S52, calculating the controllable heat exchange coefficient between each adjacent area based on the isolation distance, the isolation material parameters, and the temperature difference between adjacent areas. Specifically, a heat exchange simulation algorithm is used to calculate the controllable heat exchange coefficient, which mainly considers the combined effects of radiation, convection, and limited heat conduction. Radiative heat transfer is closely related to the emissivity of the two surfaces, the fourth power of the absolute temperature difference, and the viewing angle factor; convective heat transfer is related to the spacing, temperature difference, and airflow pattern; and the heat conduction is mainly determined by the thermal resistance of the insulation material. In one embodiment, the workshop is divided into multiple thermal equilibrium sub-regions using a region decomposition method. A local thermal network model is constructed for each adjacent sub-region, and the steady-state temperature and heat flow distribution of each sub-region are iteratively solved to obtain the equivalent heat exchange coefficient per unit temperature difference. This coefficient has an area normalization property, and its unit is usually W / (m²·K). The smaller the value, the better the insulation performance and the less heat cross-leakage. Step S53: Determine the installation position and opening setting value of the control valve based on the controllable heat exchange coefficient.

[0030] Optionally, this step also includes:

[0031] Step S531: Establish the mapping relationship between the controllable heat exchange coefficient and the opening degree of the regulating valve. Step S532: Calculate the opening setpoint of each regulating valve through the mapping relationship. Step S533: Associate the installation position of the regulating valve with the opening setpoint to form complete regulating valve setting parameters. Specifically, the regulating valves mainly include: regulating valves on the steam main pipe, combustion air regulating valves of each dryer hot air furnace, cold air inlet regulating valves of the cooler, dampers of the exhaust fan, and local forced hot air exhaust valves, etc. The selection of the installation position follows two principles: first, the heat flow direction controlled by the valve contributes the most to reducing the heat exchange coefficient of the critical section; second, it facilitates on-site operation and maintenance. The opening setpoint is determined by looking up the valve opening-heat regulation capacity mapping table obtained by pre-calculation offline, or by online model predictive control. The goal is to make the actual heat exchange volume of each adjacent section close to or slightly lower than the maximum cross heat limit allowed by the process. For example, a controllable hot air bypass valve is set between the primary dryer and the secondary dryer. When simulation results show that even with optimal isolation measures, the temperature difference between the two sections still causes approximately 18% of excess heat to flow back into the primary dryer, affecting the uniformity of particle moisture content, a bypass valve can be added to the air inlet duct of the secondary dryer to directly guide some of the preheated air to the exhaust manifold, reducing the amount of hot air entering the secondary dryer and thus effectively controlling the intensity of heat exchange between the sections. Step S6 involves performing a production process simulation to determine if the risk of particle cracking is below the cracking threshold. If it is, the current 3D workshop structure diagram and the corresponding control valve settings are determined as the optimized temperature control scheme. The production process simulation focuses on single particles or particle groups, tracking their temperature-time history from urine spraying, material mixing, granulation, drying, and cooling, and calculating the internal thermal stress distribution of the particles.

[0032] Optionally, this step also includes: Step S61, generating the temperature-time distribution for each process interval based on the setting parameters of the control valves. Based on the adjusted workshop structure, valve opening, and historical statistics of material flow rate, moisture content, urea ratio, etc., the average material temperature and temperature distribution range at key nodes such as the urine spraying tower outlet, mixer outlet, granulator outlet, primary dryer outlet, secondary dryer outlet, and cooler outlet are calculated using a segmented lumped parameter model or a one-dimensional / two-dimensional finite element model. Simultaneously, the average residence time of particles in each piece of equipment and the return ratio are considered to construct a family of temperature-time curves experienced by the particles. Step S62, calculating the changes in internal thermal stress of the particles sequentially for the urine spraying, material mixing, granulation, drying, and cooling processes. The internal thermal stress of the particles mainly originates from the difference in thermal expansion caused by uneven temperature. The calculation typically assumes the particles are spherical or approximately spherical. First, the transient temperature field distribution inside the particles is solved, and then the thermal stress components are calculated using thermoelastic theory based on parameters such as temperature field gradient, material thermal expansion coefficient, elastic modulus, and Poisson's ratio. The focus is on the alternating tensile and compressive stresses caused by the temperature difference between the particle surface and core, as well as the transient thermal shock stresses during rapid heating or cooling. In one possible implementation, the particle is divided into a three-layer lumped model: a core layer, a transition layer, and a surface layer. Different equivalent thermophysical parameters are assigned to each layer, and the temperature variation of each layer over time is numerically calculated to estimate the thermal stress difference between the layers. Step S63 involves comparing the thermal stress variation with the critical cracking stress threshold of the granular material to determine the particle cracking risk value. The critical cracking stress threshold of the granular material is typically obtained through laboratory single-particle thermal shock tests or industrial field statistics. For example, for urea-based compound fertilizer granules with a nitrogen-phosphorus-potassium ratio of 15-15-15, the typical critical cracking stress threshold during the drying stage is approximately 2.1–2.6 MPa; exceeding this value significantly increases the probability of microcrack propagation. The cracking risk value can be defined as: the cumulative percentage of time during which thermal stress exceeds the critical cracking stress threshold throughout the entire temperature history, or the ratio of maximum thermal stress to the critical cracking stress threshold, or the failure probability based on Weibull statistics, etc. Step S64 compares the particle cracking risk value with the cracking threshold and outputs the judgment result. The cracking threshold is generally determined comprehensively based on product qualification rate targets, rework costs, customer complaint tolerance, etc. For example, if the target is a finished particle crack rate of less than 0.8‰, the corresponding cracking threshold can be set between 0.15 and 0.22. When the simulation result is lower than this cracking threshold, it is considered that the current layout and control parameter combination can effectively control particle thermal damage, and it is solidified as the current optimized solution; if it is higher than the cracking threshold, it is necessary to return to step S4, adjust the weights in the optimization model, or add new control methods, and iterate again. In practical applications, after implementing this method, the finished particle crack rate of a certain device decreased from the original 1.7‰ to about 0.4‰, the secondary product rate decreased by about 2.1 percentage points, and the annual economic benefits were significant.Meanwhile, due to the improved orderly utilization of heat, the steam consumption per unit product decreased by approximately 7.4 kg / t, resulting in a significant improvement in overall energy consumption. In another implementation, the coupling calculation of the surface moisture evaporation rate and the internal moisture migration resistance can be further introduced, making the thermal stress assessment more closely reflect the actual drying kinetics. For example, when the particle surface dries rapidly to form a hard shell while the interior still contains a significant amount of free water, a large difference in internal and external shrinkage stress occurs. This "dry shell, wet core" stress is one of the important causes of cracking in urea-based compound fertilizer particles. By coupling a moisture migration model, the location and amplitude of such stress peaks can be predicted more accurately, thereby guiding the refined design of the drying segment temperature curve and airflow distribution. It should be noted that the above production process simulation is not completed in one go, but rather forms a closed-loop iterative relationship with the three-dimensional layout optimization. When the particle cracking risk does not meet the requirements, local or global layout fine-tuning and valve parameter re-optimization can be automatically triggered until the cracking threshold is met or the maximum number of iterations is reached. This closed-loop mechanism can effectively cope with thermal field drift caused by raw material fluctuations, changes in environmental temperature and humidity, and equipment aging, ensuring the long-term stability and reliability of the temperature control scheme. In another possible implementation, when a workshop simultaneously produces multiple formulations of urea-based compound fertilizer, a separate database of granular thermal stress-sensitive parameters and cracking thresholds can be maintained for each formulation. For example, high-nitrogen formulations, due to their high urea content and strong thermal sensitivity, should have a more stringent cracking threshold; while high-potassium formulations, due to the better thermal conductivity of potassium salts and their relatively dense granular structure, can have a more relaxed cracking threshold. Through adaptive formulation adjustments, differentiated and precise temperature control can be achieved during the co-production of multiple varieties in the same workshop.

[0033] If the technical solution of this application involves personal information, the product using this solution has clearly informed the user of the personal information processing rules and obtained the user's voluntary consent before processing the personal information. If sensitive personal information is involved, the user's separate consent has been obtained before processing, and the "express consent" requirement is met. For example, a clear sign is placed at the collection device such as a camera to inform the user that they have entered the collection area, and the user's voluntary entry is considered as consent; or the processing device clearly indicates the processing rules and obtains authorization through pop-up windows or by asking the user to upload information themselves. The personal information processing rules include the processor, the purpose of processing, the processing method, and the types of personal information.

[0034] Based on the embodiments of the present invention described above, and through the above description, those skilled in the art can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A complete preparation method for urea-based compound fertilizer based on intelligent gradient temperature control, characterized in that, include: S1. Collect real-time temperature and heat flow distribution data for each process area in the workshop to generate a temperature gradient distribution map; S2. Based on the temperature gradient distribution map, partition the heat transfer path to determine the spatial boundaries of the high-temperature, medium-temperature, and low-temperature zones; S3. If the spatial boundaries show disordered heat transfer characteristics, classify the temperature unstable areas to obtain the location coordinates of potential heat interference sources; S4. Input the location coordinates of potential heat interference sources into the 3D layout optimization model to obtain the adjusted 3D workshop structure diagram; S5. Extract interval isolation parameters and determine the setting parameters of the control valves through heat exchange simulation; S6. Perform production process simulation to determine whether the particle cracking risk is lower than the cracking threshold. If it is lower, determine the current 3D workshop structure diagram and the corresponding control valve setting parameters as the optimized temperature control scheme.

2. The method as described in claim 1, characterized in that, Step S1 includes: The raw data collected by the sensor network is first filtered to remove impulse noise, and then spatially interpolated to form point cloud data of temperature field and heat flow field covering the entire three-dimensional space of the workshop. Then, the temperature change rate and heat flow vector direction at each spatial point are calculated by the gradient operator, and finally a temperature gradient distribution map is generated.

3. The method as described in claim 1, characterized in that, Step S2 includes: Step S21, grouping the spatial points of the heat transfer path in the temperature gradient distribution map; Step S22, defining the three-dimensional spatial range of the high temperature zone, the medium temperature zone, and the low temperature zone according to the grouping results; Step S23, using the three-dimensional spatial range as the basis for the regional division of subsequent workshop structure adjustments.

4. The method as described in claim 3, characterized in that, Step S21 includes: A density-based spatial clustering method is used to group spatial points along heat transfer paths.

5. The method as described in claim 1, characterized in that, Step S3 includes: step S31, extracting temperature fluctuation feature vectors of temperature unstable regions; step S32, inputting the temperature fluctuation feature vectors into a support vector machine model; and step S33, outputting a set of location coordinates of potential thermal interference sources through the support vector machine model.

6. The method as described in claim 1, characterized in that, Step S4 includes: Step S41, taking the location coordinates of potential heat interference sources, the current equipment placement constraints in the workshop, and the heat transfer direction constraints as input parameters for the three-dimensional layout optimization model; Step S42, calculating feasible heat transfer paths with the goal of minimizing the ordered heat transfer path and heat interference; Step S43, adjusting the workshop equipment layout based on the feasible heat transfer paths and outputting the adjusted three-dimensional workshop structure diagram.

7. The method as described in claim 6, characterized in that, Step S42 includes: A multi-objective weighted optimization method is adopted, which linearly combines the two objectives of minimizing the orderly heat transfer path and minimizing thermal interference through weight coefficients to form a single objective function.

8. The method as described in claim 1, characterized in that, Step S5 includes: Step S51, extracting the isolation distance and isolation material parameters of adjacent functional areas from the adjusted three-dimensional workshop structure diagram; Step S52, calculating the controllable heat exchange coefficient between each adjacent area based on the isolation distance, the isolation material parameters, and the temperature difference between adjacent areas; Step S53, determining the installation position and opening setting value of the control valve based on the controllable heat exchange coefficient.

9. The method as described in claim 8, characterized in that, Step S53 includes: Step S531, establishing a mapping relationship between the controllable heat exchange coefficient and the opening degree of the control valve; Step S532, calculating the opening set value of each control valve through the mapping relationship; Step S533, associating the installation position of the control valve with the opening set value to form complete control valve setting parameters.

10. The method as described in claim 1, characterized in that, Step S6 includes: Step S61, generating the temperature-time distribution for each process interval based on the setting parameters of the control valve; Step S62, calculating the changes in thermal stress inside the particles sequentially for the processes of urine spraying, material mixing, granulation, drying, and cooling; Step S63, determining the particle cracking risk value by comparing the changes in thermal stress with the critical cracking stress threshold of the particle material; Step S64, comparing the particle cracking risk value with the cracking threshold and outputting the judgment result.

Images