A mobile water and fertilizer integrated platform suitable for green fertilizer and a control method thereof

Abstract

The application discloses a mobile water-fertilizer integrated platform suitable for green fertilizer and a control method thereof. The water-fertilizer integrated platform is integrated with a proportioning mixing tank, a feedback detection assembly, a variable-speed hose pump, a three-stage filtering assembly, a circulation backflow assembly, a reverse pulse blowdown assembly and a controller on a mobile bearing assembly. The feedback detection assembly collects viscosity, temperature, pH value and conductivity. When the concentration or pH value deviates from the target range, the controller adjusts the metering flow, stirring state and circulation backflow valve of each raw material branch, so that the mixed fertilizer solution returns to the proportioning mixing tank for re-compounding. Qualified fertilizer solution is sequentially output for application after passing through a 80-120 mu m screen, a magnetic particle separation device and a 55 mu m automatic backwashing laminated filter assembly. The controller performs blockage diagnosis according to the pressure and flow at each stage, and when the trigger condition is met, the reverse pulse solenoid valve and the blowdown valve are linked to complete automatic flushing, thereby improving the application accuracy and reducing the risk of blockage.

Description

Technical Field

[0001] This invention belongs to the technical field of integrated water and fertilizer application and intelligent irrigation and fertilization equipment in agriculture, specifically relating to a mobile integrated water and fertilizer platform adapted to green fertilizers and its control method. Background Technology

[0002] Integrated water and fertilizer technology combines irrigation and fertilization processes through pumps, valves, pipelines, filters, and controllers, enabling quantitative delivery and field application of fertilizer solutions. Existing integrated water and fertilizer systems typically treat ordinary water-soluble fertilizers or clarified fertilizer solutions, and are mainly designed around water supply filtration, fixed-ratio fertilizer injection, and pipeline pressure control.

[0003] Humic acid fertilizers, liquid microbial fertilizers, and microbial inoculants, among other green and efficient fertilizers, contain colloids, inoculant clusters, or suspended particles. Their viscosity, pH value, temperature, concentration, and particle sedimentation characteristics differ significantly from those of conventional fertilizer solutions. When multiple green and efficient fertilizers are mixed with conventional fertilizers, uneven dissolution, agglomeration, flocculation, precipitation, or crystallization can easily occur, leading to mixing ratio errors, fluctuations in fertilizer injection flow, and blockages in filters, bends, and field pipelines.

[0004] Existing equipment typically relies on manual experience or fixed pump speeds to adjust the ratio of concentrate to water, making it difficult to provide real-time feedback based on changes in fertilizer solution viscosity, pH, and temperature. Filters are mostly single-stage or passive filtration structures, unable to identify blockage locations based on pressure differences and flow rate changes before and after the filter, and also unable to automatically perform reverse pulse flushing during field operations. Furthermore, existing technologies often lack clear distinctions between return pipelines for circulation mixing, backwash pipelines for filtration cleaning, and sewage discharge pipelines for removing contaminants, easily leading to inconsistencies between structural specifications and control logic. Summary of the Invention

[0005] The purpose of this invention is to provide a mobile fertigation platform and its control method adapted to green fertilizers. It can achieve precise multi-source fertilizer solution ratio, online status detection, circulation reflux homogenization, three-stage gradient filtration, pressure-flow coupling blockage diagnosis, and automatic reverse pulse flushing. It can effectively solve the problems of unstable ratio when green synergistic fertilizers are used together with conventional fertilizers, failure to provide timely feedback on pH status and viscosity changes, difficulty in locating filter blockages, and unclear functions of backwashing pipelines and circulation reflux pipelines.

[0006] To achieve the above objectives, the present invention provides a mobile fertigation platform adapted to green fertilizers, comprising a mobile support component, a controller for centralized control of the fertigation platform, and a proportioning and mixing component, a quantitative conveying component, and a three-stage filtration component connected sequentially along the material conveying direction via pipelines on the mobile support component. The three-stage filtration component is equipped with pressure detection devices and flow detection devices for acquiring the filtration pressure and flow rate at each stage. The proportioning and mixing component includes a proportioning and mixing tank, which is equipped with a stirrer. The inlet of the proportioning and mixing tank is connected to the stock solution component and the multi-source metering and dispensing component via a manifold. A feedback detection component is installed on the pipeline connecting the proportioning and mixing component and the quantitative delivery component. The feedback detection component includes a viscosity detection unit, a temperature sensor, a pH sensor, and a conductivity sensor. The pipeline connecting the quantitative conveying component and the three-stage filtration component is equipped with a circulation return pipe that is connected to the return port of the proportioning mixing tank, and the circulation return pipe is equipped with a circulation return valve; The three-stage filtration assembly includes a first-stage screen filter, a second-stage magnetic particle separator, and a third-stage automatic backwashing disc filter assembly connected sequentially along the material conveying direction. The third-stage automatic backwashing disc filter assembly includes a filter housing. The upper part of the filter housing forms an inlet chamber and a clear liquid chamber through the disc filter elements. The lower part of the filter housing has a dirt collection area communicating with the inlet chamber. The outlet end of the dirt collection area is connected to a drain pipe, and a drain valve is installed on the drain pipe. The clear liquid chamber is connected to a multi-channel output control assembly through an output pipe and to a backwashing liquid supply assembly through a backwashing pipe, and a reverse pulse solenoid valve is installed on the backwashing pipe.

[0007] As a further aspect of the present invention: the raw liquid component includes a raw liquid tank and a clean water tank. The outlet end of the raw liquid tank is connected to the manifold via a raw liquid outlet pipe, and the outlet end of the clean water tank is connected to the manifold via a clean water outlet pipe. The raw liquid tank is provided with a light-shielding and heat-insulating phase change cold storage component, which includes at least one of a light-shielding shell, a heat-insulating interlayer, and a replaceable phase change cold storage module. The multi-source metering and distribution assembly includes several multi-source distribution branches. Each multi-source distribution branch is sequentially connected to a branch metering pump, a distribution electric proportional valve, a branch flow sensor, and a branch check valve. All multi-source distribution branches are connected to the manifold through a common distribution manifold.

[0008] As a further aspect of the present invention: the pressure detection element includes a filter column inlet pressure sensor located before the inlet of the first-stage screen filter, and a first pressure sensor, a second pressure sensor, and a third pressure sensor located at the outlet of the first-stage screen filter, the outlet of the second-stage magnetic particle separator, and the outlet of the third-stage automatic backwashing disc filter assembly, respectively; the flow detection element includes a first flow sensor, a second flow sensor, and a third flow sensor located at the outlet of the first-stage screen filter, the outlet of the second-stage magnetic particle separator, and the outlet of the third-stage automatic backwashing disc filter assembly, respectively.

[0009] As a further embodiment of the present invention: the multi-channel output control component includes an output manifold and several field fertilization branches. The output manifold is connected to the output pipe, and all field fertilization branches are connected in parallel to the output manifold. Each field fertilization branch is equipped with an output electric proportional valve and a branch flow sensor.

[0010] As a further aspect of the present invention: the backwashing liquid supply assembly includes a filter isolation valve, a backwashing check valve, and a backwashing pump connected in sequence along the direction away from the reverse pulse solenoid valve.

[0011] To achieve the above-mentioned objectives, the present invention also provides a control method for a mobile fertigation platform adapted to green fertilizers, comprising the following steps: S1, move the integrated water and fertilizer platform to the work site, connect the multi-channel output control component, and initialize the target fertilizer solution ratio, target concentration, target pH range, allowable error and sampling cycle; S2 controls the original liquid component and the multi-source metering and application component to send the corresponding materials into the mixing tank for mixing to obtain fertilizer solution; S3 collects fertilizer solution viscosity, temperature and pH value through feedback detection components, filters and compensates for temperature in the collected data, and estimates the current fertilizer solution concentration; S4, determine whether the current fertilizer solution concentration and pH value are within the corresponding target range; if not, turn off the multi-channel output control component, turn on the circulation reflux valve, adjust one or more of the following: the metering flow rate of the original solution component and the multi-source metering and application component, the stirring frequency of the mixing tank and the mixing time, and return to step S3 after the delay stabilizes. S5, when the current fertilizer solution concentration and pH value are within the corresponding target range, close the circulation reflux valve and control the quantitative delivery component to make the fertilizer solution pass through the first-stage screen filter, the second-stage magnetic particle separation device and the third-stage automatic backwashing disc filter component in sequence before being output; S6 collects the filtration pressure and flow rate of each stage of the three-stage filtration assembly, and determines the risk of clogging based on the coupling relationship between pressure and flow rate; S7, when the blockage risk meets the preset triggering conditions, reduce or stop the quantitative delivery component, close the circulation return valve and the multi-channel output control component, open the reverse pulse solenoid valve, the drain valve and the backwash liquid supply component, so that the flushing medium passes through the clear liquid chamber of the third-stage automatic backwash disc filter component in the reverse direction and discharges the pollutants through the drain pipe. S8: After rinsing, re-collect the pressure and flow of each stage of the filter. When the pressure and flow return to the preset range, resume normal output of step S5; otherwise, repeat rinsing, alarm, or shutdown.

[0012] As a further aspect of the present invention: S3 also includes collecting the conductivity EC and the current fertilizer solution concentration. Based on the temperature-compensated viscosity μ, pH value, temperature T, and conductivity EC, the following parameters are calculated using regression models, lookup table models, interpolation models, or calibrated machine learning models: =f(μ,pH,T,EC).

[0013] As a further aspect of the present invention: In S6, the inlet pressure of the front-end filter column of the three-stage filtration assembly is denoted as P0, and the outlet pressures of the first-stage screen filter, the second-stage magnetic particle separator, and the third-stage automatic backwashing disc filter assembly are denoted as P1, P2, and P3 respectively, and the corresponding outlet flow rates are denoted as Q1, Q2, and Q3 respectively; the pressure differences between each stage are ΔP1=P0-P1, ΔP2=P1-P2, and ΔP3=P2-P3 respectively; the controller according to... Determine the clean baseline pressure differential for the i-th stage of filtration, and based on... Determine the pressure-flow residual, where i is the filter stage number, and ai and bi are coefficients obtained from equipment calibration, historical operating data, or computational fluid dynamics analysis.

[0014] As a further aspect of the present invention: S6 uses one or more of the following methods to identify congestion trends: sliding window mean, rate of change, cumulative sum control chart, threshold combination, or machine learning classification model.

[0015] As a further aspect of the present invention: when normal filtration is performed and the recirculation valve, reverse pulse solenoid valve, and drain valve are all closed, Q1≈Q2≈Q3 is used as the flow continuity verification condition; when the recirculation valve, reverse pulse solenoid valve, drain valve, or multi-channel output control component is opened, the corresponding recirculation flow rate, flushing flow rate, drain flow rate, or branch flow rate is added to the mass conservation relationship, and the diagnostic model of the corresponding operating state is switched.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. By forming a closed loop of "detection-decision-execution-retest" through multi-source metering and application components, feedback detection components and controllers, it can adjust the ratio of clean water to different fertilizer solution branches according to changes in viscosity, pH value, temperature and conductivity, reducing the ratio fluctuation caused by fixed pump speed or manual experience adjustment.

[0017] 2. The mixed fertilizer solution that has not reached the target concentration or target pH range is returned to the mixing tank for circulation and homogenization through the circulation return pipe and circulation return valve. The circulation return pipe is independent of the reverse pulse flushing pipe to prevent flushing contaminants from flowing back to the mixing tank.

[0018] 3. A three-stage gradient filtration system is formed by a first-stage screen filter, a second-stage magnetic particle separation device, and a third-stage automatic backwashing disc filter assembly. This system separately treats large suspended solids, ferromagnetic particles, and fine particles such as humic acid colloids and bacterial agent clusters, thereby reducing the concentrated load on a single-stage filter.

[0019] 4. By constructing pressure-flow coupling diagnostics through P0~P3 and Q1~Q3, it is possible to distinguish between changes in fertilizer solution viscosity, progressive filter blockage, and sensor malfunctions, and to automatically flush the reverse pulse solenoid valve and drain valve when the trigger conditions are met.

[0020] 5. Through several multi-source metering and application components and multi-channel output control components, it can adapt to the dynamic application needs of different fertilizer combinations, different fields and different crop growth stages; the mobile bearing, modular layout and optional shading, heat preservation and phase change cold storage structure are conducive to mobile field operations and environmental adaptation. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the structure of the mobile water and fertilizer integration platform adapted to green fertilizers according to the present invention. Figure 2 This is a schematic diagram of the pipeline under normal filtration, circulation reflux, and reverse pulse flushing states of the present invention; Figure 3 This is a detailed structural diagram of the first-stage screen filter of the present invention; Figure 4 This is a detailed structural diagram of the second-stage magnetic particle separation device of the present invention; Figure 5 This is a detailed structural diagram of the third-stage automatic backwashing disc filter assembly of the present invention; Figure 6 This is a diagram illustrating the structure and control relationship of the multi-source metering application and pH feedback regulation of the present invention; Figure 7 This is a schematic diagram of the pressure-flow coupling blockage diagnosis and flushing triggering of the present invention; Figure 8 This is a flowchart illustrating the control method for a mobile integrated water and fertilizer platform adapted to green fertilizers according to the present invention.

[0022] In the diagram: 1. Mobile support assembly; 2. Raw material tank; 2.1 Raw material outlet pipe; 3. Clean water tank; 3.1 Clean water outlet pipe; 4. Stirrer; 5. Proportioning and mixing tank; 5.1 Manifold; 6. Quantitative delivery assembly; 6.1 Adjustable speed hose pump; 6.2 Check valve; 7. First-stage screen filter; 7.1 First pressure sensor; 7.2 First flow sensor; 7.3 Stainless steel screen; 7.4 First-stage discharge port; 8. Second-stage magnetic particle separator; 8.1 Second pressure sensor; 8.2 Second flow sensor. 8.3 Separation Housing; 8.4 Magnetic Collection Component; 8.5 Magnetic Impurity Collection Structure; 9. Feedback Detection Component; 9.1 Viscosity Detection Unit; 9.2 Temperature Sensor; 9.3 pH Sensor; 9.4 Conductivity Sensor; 10. Filter Column Inlet Pressure Sensor; 11. Third-Stage Automatic Backwashing Disc Filter Assembly; 11.1 Filter Housing; 11.2 Liquid Inlet Chamber; 11.3 Disc Filter Element; 11.4 Clear Liquid Chamber; 11.5 Third Pressure Sensor; 11.6 Third Flow Sensor; 12. 13. Output pipe; 14. Reverse pulse solenoid valve; 15. Sewage pipe; 16. Sewage valve; 17. Sewage collection area; 18. Circulation return pipe; 19. Circulation return valve; 20. Controller; 20. Backwash supply assembly; 20.1. Backwash pipe; 20.2. Backwash pump; 20.3. Backwash check valve; 20.4. Filter isolation valve; 21. Multi-source metering and application assembly; 21.1-i. Multi-source application branch; 21.2-i. Branch metering pump; 21.3-i. Application electric proportional valve; 21.4-i. Branch flow sensor. 21.5-i, Branch check valve; 21.6, Common application manifold; 22, Multi-channel output control assembly; 22.1, Output manifold; 22.2-i, Output electric proportional valve; 22.3-i, Branch flow sensor; 22.4-i, Field fertilization branch; 24, Shading and heat preservation phase change cold storage assembly. In the attached figures, the subscript i only indicates the ordinal number of the same type of branch and is not included in the calculation of the formulas in this manual. In the formulas of this manual, i represents the filter stage number, the material branch number is represented by j, and the field fertilization branch number is represented by n. Detailed Implementation

[0023] The invention will now be further described with reference to the accompanying drawings.

[0024] like Figure 1As shown, a mobile fertigation platform adapted for green fertilizers includes a mobile load-bearing component 1. The mobile load-bearing component 1 can be a single-axle frame with a traction structure, a mobile frame with casters, or a modular chassis mounted on an agricultural transport platform. The raw material tank 2, clean water tank 3, proportioning and mixing tank 5, quantitative conveying component 6, three-stage filtration component, backwashing liquid supply component 20, and controller 19 are all fixed to the mobile load-bearing component 1 via vibration-damping supports or detachable connectors. The platform adopts a compact, modular layout, and the overall design weight is preferably no more than 300 kg; this weight value is a design constraint, and the specific weight is determined by the tank volume, pump and valve specifications, and load-bearing structure.

[0025] The stock solution tank 2 is used to store humic acid stock solution or other green synergistic fertilizer stock solutions; the clear water tank 3 is used for proportioning and dilution, pipeline pre-filling, and reverse pulse flushing. Both the stock solution tank 2 and the clear water tank 3 are equipped with a filling port, a level detector, and a bottom drain. The proportioning and mixing tank 5 is equipped with a reflux port and a low-shear agitator 4 to reduce sedimentation and agglomeration while avoiding unnecessary high-shear damage to the active bacterial agent.

[0026] The multi-source metering and application component 21 is equipped with four to eight multi-source application branches 21.1-i according to the specific fertilizer combination. The branch metering pump 21.2-i of each branch can be a diaphragm metering pump, peristaltic pump or other corrosion-resistant adjustable speed pump; the application electric proportional valve 21.3-i, the branch flow sensor 21.4-i and the branch check valve 21.5-i of each branch are used to form mutually independent closed-loop metering to prevent cross-flow between different raw liquid branches.

[0027] like Figure 2 As shown, the feedback detection component 9 is located between the outlet of the mixing tank 5 and the inlet of the adjustable-speed hose pump 6.1. This arrangement allows the mixed fertilizer solution to have its viscosity, temperature, pH value, and conductivity detected before entering the adjustable-speed hose pump 6.1 and the three-stage filtration assembly. It also ensures that the fertilizer solution in the reflux state passes through the feedback detection component 9 again in each circulation cycle.

[0028] The recirculation return pipe 17 extends from the outlet of the adjustable speed flexible hose pump 6.1 to the inlet of the first-stage screen filter 7 and reconnects to the proportioning mixing tank 5. A recirculation return valve 18 is installed on the recirculation return pipe 17. The recirculation return valve 18 is only used to switch the recirculation return path of the mixed fertilizer solution and does not perform pH detection, acid / alkali addition, or reverse pulse flushing functions. Its function is to allow the mixed fertilizer solution to return to the proportioning mixing tank 5 for recirculation mixing when the concentration, viscosity, or pH value of the mixed fertilizer solution has not reached the target range, the sensor is in a stable sampling stage, the output is temporarily closed, or the pump outlet needs to be depressurized.

[0029] Under normal fertilization conditions, the recirculation valve 18 is closed or kept open to meet the minimum circulation flow rate, allowing qualified mixed fertilizer solution to enter the first-stage screen filter 7. Under circulation adjustment conditions, the controller 19 closes the normal output branch and opens the recirculation valve 18. Under reverse pulse flushing conditions, the recirculation valve 18 remains closed, thereby isolating the recirculation pipe 17 from the backwash pipe 20.1, the sludge collection area 16, and the drain pipe 14, preventing flushing contaminants from entering the proportioning mixing tank 5.

[0030] The pH sensor 9.3 in the feedback detection component 9 is only responsible for detecting the pH value of the mixed fertilizer solution. The pH feedback adjustment module in the controller 19 receives the detection result from the pH sensor 9.3 and adjusts the metering flow rate of the existing multi-source application branch, the stirring frequency of the stirrer 4, the mixing time, and the opening of the circulation reflux valve 18 to re-proportion and homogenize the mixed fertilizer solution. In this embodiment, no separate acid or alkaline adjustment solution storage tank, metering pump, and dosing pipeline are provided; the pH feedback adjustment module is only implemented as a software function in the controller 19.

[0031] To facilitate understanding of the meaning of each formula in this manual, the symbols used in the formulas are explained uniformly first; throughout the entire manual, the same symbol has only the following unique meaning.

[0032] Sequence and time-related symbols: i is the filter stage sequence number, i=1, 2, 3, corresponding to the first-stage screen filter 7, the second-stage magnetic particle separator 8, and the third-stage automatic backwashing disc filter assembly 11 respectively (i in the attached diagrams 21.1-i, 22.2-i, etc., is only the ordinal number of the same type of branch and is not involved in the formula calculation); j is the material branch sequence number; n is the field fertilization branch sequence number; q is the calibration sample sequence number; t is time; dt is a small time interval; τ is the duration of the operation period within a configuration cycle.

[0033] Fertilizer solution ratio and concentration symbols: pH is the pH value of the mixed fertilizer solution; pHmin and pHmax are the lower and upper limits of the target pH range, respectively; e is a general symbol for deviation, and its subscript indicates the controlled quantity corresponding to the deviation, where e+pH is the upper pH deviation, e-pH is the lower pH deviation, epH is the comprehensive pH deviation, eC is the concentration error, and ei is the pressure-flow residual of the i-th stage filtration; Qj, ρj, and cj are the volumetric flow rate, material density, and target component mass fraction of the j-th material branch, respectively; For the target component mass flow rate, The total mixed mass flow rate is represented by a dot above the symbol, indicating that the quantity is the mass per unit time; Cbal is the physical reference value for concentration based on the flow rate. This represents the estimated current fertilizer concentration; the "^" above the symbol indicates an estimated or predicted value calculated by the model; Ctarget is the target concentration; f is the concentration estimation function; θ is the parameter vector of the concentration estimation function. θ0, θ1, θ2, θ3, and θ4 are the regression coefficients of the linear calibration model; xq is the feature vector of the q-th calibration sample; ω is the regularization coefficient; and EC is the electrical conductivity of the mixed fertilizer solution.

[0034] Viscosity and temperature symbols: μ is the dynamic viscosity of the mixed fertilizer solution, μmea is the measured viscosity, μT is the viscosity after temperature compensation, μ0 is the vacuum permeability; T is the temperature of the mixed fertilizer solution, Tref is the reference temperature for calibration, Tmax is the upper limit of the target temperature, T0 is the initial temperature of the fertilizer solution, Ta is the ambient temperature, T∞ is the equilibrium temperature without cold storage, Tm is the phase change temperature, Tpcm,0 is the initial temperature of the phase change material, and τT is the temperature response time constant.

[0035] Pressure, flow rate, and filtration symbols: P represents pressure, P0 is the inlet pressure of the filter column, P1, P2, and P3 are the outlet pressures of the first-stage screen filter 7, the second-stage magnetic particle separator 8, and the third-stage automatic backwashing disc filter assembly 11, respectively; ΔPi is the measured pressure difference between the inlet and outlet of the i-th stage filter assembly. ρ is the clean baseline pressure difference of the i-th stage filter element; Q is the system volumetric flow rate during normal series filtration; Qi is the outlet volumetric flow rate of the i-th stage filter element; V12 and V23 are the liquid volumes in the control bodies of the pipe sections between adjacent flow measurement points; ρ is the density of the mixed fertilizer solution; g is the acceleration due to gravity; v is the average velocity of the pressure measurement section; vf,i is the apparent velocity of the i-th stage filter medium; z is the elevation of the measurement point; α is the kinetic energy correction coefficient; hL,i is the head loss generated by the i-th stage filter unit and its connecting pipes; x is the coordinate along the material flow direction; ai and bi are the viscous resistance equivalent coefficient and inertial resistance equivalent coefficient of the i-th stage filter element, respectively; aΣ and bΣ are the sum of the viscous resistance equivalent coefficients of the three stages and the sum of the inertial resistance equivalent coefficients of the three stages, respectively. and ; ki is the equivalent permeability of the i-th stage filter medium, δi is the thickness of the i-th stage filter medium, Af,i is the effective filtration area of ​​the i-th stage filter medium, βi is the inertial resistance coefficient of the i-th stage filter medium; λi is the friction coefficient of the i-th stage connecting pipe section, Lp,i, Di and Ap,i are the length, diameter and flow area of ​​the i-th stage connecting pipe section, respectively, ζi is the local resistance coefficient; As,0 is the total area of ​​the screen, φs is the initial opening ratio of the screen, σs is the proportion of the area occupied by the intercepted material in the screen holes, As,eff is the effective flow area of ​​the screen, Cd,s is the flow coefficient of the screen holes; Fm is the magnetic force on the magnetic particles, Vp is the volume of the magnetic particles, Δχ is the difference in magnetic susceptibility between the magnetic particles and the mixed fertilizer solution, and B is the magnetic induction intensity.

[0036] Symbols for shading, heat preservation, and phase change cold storage: V is the effective volume of fertilizer solution in tank 2; mf is the mass of fertilizer solution in the tank; cp is the specific heat capacity at constant pressure of the mixed fertilizer solution; dEf is the change in sensible heat of the fertilizer solution; U is the overall heat transfer coefficient of the tank; Ah is the effective heat exchange area of ​​the tank; γ is the effective absorption coefficient of the fertilizer solution to solar radiation after modification by the shading structure; I is the solar irradiance; Arad is the effective projected area exposed to solar radiation; ηw is the heat conversion factor for internal heat input; Waux is the operating power of the stirrer, pump, and other auxiliary components; dEpcm / dt is the instantaneous heat absorption of the phase change cold storage module. Power; E0 is the initial sensible heat of overheating; Ereq is the required cold storage energy within the configuration cycle; Ks is the cold storage safety factor; ΔHm is the latent heat of phase change; cs and cl are the solid-phase heat capacity before phase change and the liquid-phase heat capacity after phase change, respectively; ueff is the available cold storage capacity per unit mass of phase change material in the target temperature range; ηpcm is the cold storage utilization efficiency; mpcm is the configuration mass of phase change material; mmod is the mass of phase change material in a single phase change cold storage module; Epcm,usable is the available cold storage capacity that phase change material can provide; N is the configuration number of phase change cold storage modules.

[0037] Operators: Σ represents summation; ∫ represents integration; Represents partial derivatives; The gradient operator is represented by ; max() represents taking the maximum value among the terms within the parentheses; ⌈·⌉ represents rounding up; ‖·‖ represents the second norm of a vector; exp() represents an exponential function with the natural logarithm as the base; the symbol “≈” indicates approximate equality within the measurement error range.

[0038] If the target pH range is set as [pHmin, pHmax], where pHmin is the lower limit of the target pH range and pHmax is the upper limit of the target pH range, then the pH deviation can be directly derived from the boundaries of the target range: ; ; ; Wherein, pH is the pH value of the mixed fertilizer solution detected by pH sensor 9.3 and filtered; e is a general symbol for the deviation, and its subscript indicates the controlled quantity corresponding to the deviation; e+pH is the upper pH deviation, which represents the value of the detected pH value that is higher than the target upper limit pHmax, and is zero when the detected value is not higher than the target upper limit; e-pH is the lower pH deviation, which represents the value of the detected pH value that is lower than the target lower limit pHmin, and is zero when the detected value is not lower than the target lower limit; epH is the comprehensive pH deviation, which is equal to the upper pH deviation minus the lower pH deviation; max(·, 0) means taking the larger of the two items in parentheses, that is, taking zero when the first item in parentheses is negative.

[0039] When ePH is zero, the pH value is within the target range; when ePH is positive, the detected value is higher than the target upper limit; when ePH is negative, the detected value is lower than the target lower limit. Controller 19 adjusts the metering flow rates of the clean water, humic acid, liquid microbial fertilizer, and conventional fertilizer branches according to the sign and amplitude of ePH, and links the stirrer 4 and the circulation return valve 18; after a delay and stabilization, it resamples, thus forming a closed-loop regulation between the pH sensor 9.3 and the actuator.

[0040] To illustrate the origin of the concentration estimation function f, let the volumetric flow rate of the j-th material branch within a sampling period be Qj, the density be ρj, and the mass fraction of the target component be cj, where j is the branch number. By conserving component mass, the mass flow rate of the target component brought into each branch is... and total mixed mass flow rate (The dot above the symbol indicates that the quantity is the mass per unit time, i.e., the mass flow rate): ; Where Σj represents the summation over all connected material branches; The target component mass flow rate is the mass of each material branch brought into the mixing tank 5 per unit time. The total mass of material entering the mixing tank 5 per unit time is the total mass flow rate of the mixture.

[0041] When low-shear stirring is carried out in mixing tank 5, the sampling period is short enough, and the outlet state is approximately stable, the concentration physical benchmark Cbal based on the flow rate, i.e., the benchmark value of the target component mass fraction of the mixed fertilizer solution calculated from the volumetric flow rate and density of each branch, is as follows: ; Cbal is used to prepare calibration samples and verify online estimation results without requiring linear mixing of the pH values ​​of different materials. For viscosity detection unit 9.1, a first-order temperature compensation is performed on the measured viscosity μmea near the reference temperature Tref to obtain the temperature-compensated viscosity μT: ; Where μ is the dynamic viscosity of the mixed fertilizer solution; μmea is the measured viscosity of the viscosity detection unit 9.1; μT is the viscosity after temperature compensation; Tref is the reference temperature for calibration; and T is the temperature of the mixed fertilizer solution detected by the temperature sensor 9.2. The viscosity-temperature coefficient (VTC) is the rate of change of the mixed fertilizer solution viscosity with temperature at a reference temperature. This partial derivative is obtained from temperature calibration tests of the same fertilizer combination. If no online viscosity sensor is installed, the equivalent μT is obtained from the calibration relationship between pump speed, pipeline pressure, flow rate, and temperature. The feature vector xq=[1, μT, pH, T, EC] of each calibration sample is used as input, where q is the calibration sample number, and the components of the feature vector are, in order, the constant term 1, the temperature-compensated viscosity μT, the pH value, the temperature T, and the conductivity EC. When using least squares calibration with a regularization term, the model parameters can be expressed as: ; Where Σq represents the summation over all calibration samples; Cbal,q is the concentration benchmark value of the q-th calibration sample calculated according to the above concentration physical benchmark; xq is the feature vector of the q-th calibration sample; f is the concentration estimation function; θ is the parameter vector of the concentration estimation function; θ is the optimal parameter vector that minimizes the objective function within the curly braces; arg min represents the parameter that minimizes the objective function; ω is the regularization coefficient, used to suppress overfitting caused by excessively large parameters; ‖θ‖² is the sum of squares of the components of the parameter vector θ, i.e., the square of its L2 norm.

[0042] In one embodiment employing a linear calibration model, the expansion yields: ; in, This represents the estimated concentration of the current fertilizer solution. The "^" above the symbol indicates that this value is an estimate calculated by the model. θ0 is a constant term, and θ1, θ2, θ3, and θ4 are the regression coefficients corresponding to the viscosity μT, pH value, temperature T, and conductivity EC after temperature compensation, respectively. Their values ​​are obtained from the optimal parameter vector obtained above. Confirmed; EC is the conductivity of the mixed fertilizer solution detected by conductivity sensor 9.4.

[0043] When using lookup tables, interpolation, or machine learning models, the concentration estimation function f is still determined using the same set of calibration samples consisting of Cbal, μT, pH, T, and EC. Therefore, the current general expression for fertilizer solution concentration is: ; In other parts of this specification and in the claims, the viscosity used in concentration estimation is represented by μ, and its value is the viscosity μT after temperature compensation mentioned above.

[0044] Controller 19 further calculates the concentration error eC: ; Where eC represents the concentration error, and Ctarget is the target concentration initialized in step S1; eC and ePH are used together to determine the circulation reflux, branch metering flow, stirring status, and normal output permittability. The function parameters are determined by calibration experiments under different fertilizer types, target concentrations, temperatures, and mixing conditions.

[0045] like Figure 3 As shown, the first-stage screen filter 7 includes a housing, a removable stainless steel screen 7.3, and a primary drain outlet 7.4. The pore size of the stainless steel screen 7.3 is preferably 80–120 μm, used to intercept large suspended particles. A filter column inlet pressure sensor 10 is installed before the inlet of the first-stage screen filter 7, and a first pressure sensor 7.1 and a first flow sensor 7.2 are installed at its outlet to obtain the pressure before and after the first-stage filtration and the outlet flow rate.

[0046] like Figure 4 As shown, the second-stage magnetic particle separation device 8 includes a separation housing 8.3 and a detachable magnetic trapping component 8.4. A magnetic impurity collection structure 8.5 is connected below the separation housing 8.3. The magnetic trapping component 8.4 can be a magnetic rod, a magnetic sleeve, or a magnetic component arranged along the flow channel, and it is used to trap ferrous magnetic particles. The second-stage magnetic particle separation device 8 is not used to replace the filtration of non-magnetic suspended matter by a screen or disc filter. After the magnetic trapping component 8.4 reaches a preset load, it can be removed from the separation housing 8.3 and disposed of centrally.

[0047] like Figure 5 As shown, the disc filter element 11.3 of the third-stage automatic backwash disc filter assembly 11 preferably has a filtration precision of 55 μm to trap humic acid colloids, bacterial agent clusters, and fine particles. Under normal filtration conditions, the mixed fertilizer solution enters the clear liquid chamber 11.4 from the inlet chamber 11.2 through the disc filter element 11.3, and then enters the multi-channel output control assembly 22 via the outlet pipe 12. The sludge collection area 16 below the disc filter element 11.3 is used to collect sediment from normal filtration and stripped material from backwashing.

[0048] Under normal series filtration, with the recirculation valve 18 closed, the reverse pulse solenoid valve 13 closed, the drain valve 15 closed, and the output branch stable, first write the volume conservation for the pipe section between adjacent flow measurement points. Let the liquid volume in the control body of the pipe section between the first flow sensor 7.2 and the second flow sensor 8.2 (i.e., between measurement point P1 and P2) be V12, and the liquid volume in the control body of the pipe section between the second flow sensor 8.2 and the third flow sensor 11.6 (i.e., between measurement point P2 and P3) be V23, and t be time, then: ; When a rigid pipeline is full and in a steady-state, normal filtration state, the volume of the liquid inside is controlled to remain constant over time, i.e.: ; This leads to the theoretical relationship between the three series flow measurement points: ; Where Q represents the system volumetric flow rate through the three-stage filter assembly under normal series filtration conditions. Considering flow pulsation, sensor resolution, and sampling asynchrony, it is written in engineering as: ; In the formula, the symbol “≈” indicates that they are approximately equal within the above measurement error range. When the recirculation valve 18, the reverse pulse solenoid valve 13, the drain valve 15, or multiple field fertilization branches are in the open state, the recirculation flow rate, flushing flow rate, drain flow rate, and branch flow rate should be added to the mass conservation formula of the corresponding control body, and Q1≈Q2≈Q3 should no longer be directly used as the fault criterion.

[0049] For the i-th stage filter unit, where i is the filter stage number (i = 1, 2, 3), Bernoulli's mechanical energy equation is written between the inlet pressure measurement section i−1 and the outlet pressure measurement section i: ; Where P is the pressure, P(i-1) and Pi are the pressures at the inlet and outlet pressure measuring sections of the i-th stage filtration unit, respectively; ρ is the density of the mixed fertilizer solution; g is the acceleration due to gravity; v is the average flow velocity at the pressure measuring section, v(i-1) and vi are the average flow velocities at the inlet and outlet pressure measuring sections, respectively; z is the elevation of the measuring point, z(i-1) and zi are the elevations of the inlet and outlet measuring points, respectively; α is the kinetic energy correction coefficient, α(i-1) and αi are the kinetic energy correction coefficients at the inlet and outlet pressure measuring sections, respectively; hL,i is the head loss generated by the i-th stage filtration unit and its connecting pipes. When adjacent measuring points are set on pipe sections of equal height and diameter, the velocity term and the elevation term cancel each other out, thus obtaining: ; In the formula, ΔPi is the measured pressure difference between the inlet and outlet of the i-th stage filter component.

[0050] The pressure loss within the filter medium is composed of both viscous seepage loss and inertial loss. Using the Darcy-Forchheimer relation, integrating over the thickness δi of the i-th stage filter medium, and incorporating the local losses from the shell, elbows, constrictions, and connecting pipe sections into equivalent coefficients, it can be written as: ; In the formula, P is the pressure of the liquid in the filter medium, x is the coordinate along the material flow direction, dP / dx is the rate of change of pressure along the flow direction; vf,i is the apparent flow rate of the i-th stage filter medium, i.e., the filtration rate, which is equal to the ratio of the outlet volumetric flow rate Qi of that stage to the effective filtration area Af,i; δi is the thickness of the i-th stage filter medium; μ is the dynamic viscosity of the mixed fertilizer solution; and ρ is the density of the mixed fertilizer solution.

[0051] Integrating the above equation and adding the losses from the connecting pipelines, we obtain the pressure-flow equivalence relationship for each stage of the filter assembly: ; Where ai is the viscous resistance equivalent coefficient of the i-th stage filter element, and bi is the inertial resistance equivalent coefficient of the i-th stage filter element. Their representative expressions can be written as follows: ; ; In the formula, ki is the equivalent permeability of the i-th stage filter medium, Af,i is the effective filtration area of ​​the i-th stage filter medium, βi is the inertial resistance coefficient of the i-th stage filter medium, λi is the friction resistance coefficient of the i-th stage connecting pipe section, Lp,i, Di, and Ap,i are the length, diameter, and flow area of ​​the i-th stage connecting pipe section, respectively, ζi is the local resistance coefficient at the inlet, outlet, elbow, and constriction / expansion points of this stage, and Σζi is the sum of the above local resistance coefficients. aiQi represents viscous loss, biQi 2 This represents inertia, friction loss, and local losses. Under clean conditions, ai and bi are obtained through calibration or fluid dynamics analysis. After contaminant accumulation, the equivalent resistance change is reflected by the pressure-flow residual.

[0052] For the first-stage screen filter 7, let the total screen area be As,0, the initial aperture ratio be φs, and the proportion of the screen area occupied by the trapped material be σs, where the subscript s represents the screen. Then, the effective flow area As,eff of the screen is obtained from the geometric aperture relationship: ; Where As,0 is the total area of ​​the screen, φs is the initial opening ratio of the screen, σs is the proportion of the screen opening occupied by the trapped material, and As,eff is the effective flow area of ​​the screen. The screen openings are equivalent to parallel small holes, and the outflow relationship of the small holes is adopted: ; In the formula, Cd,s is the flow coefficient of the sieve aperture, i.e., the orifice outflow coefficient, which is determined by the sieve structure and calibration test; P0 is the inlet pressure of the filter column; P1 is the outlet pressure of the first-stage sieve filter 7; ρ is the density of the mixed fertilizer solution. The pressure drop of the first-stage sieve filter 7 is obtained by rearranging: ; As can be seen from this formula, when the flow rate and fertilizer solution density remain constant, a decrease in the effective opening area of ​​the screen will increase the pressure drop in the first stage. If the effects of screen thickness and viscosity cannot be ignored, the aforementioned general formula P0-P1=a1Q1+b1Q1² is adopted. The two models correspond to the local outflow model of the screen opening and the equivalent resistance model of the filter column, respectively, and the pressure loss of the same filter medium is not repeatedly accumulated.

[0053] For the second-stage magnetic particle separation device 8, the magnetic field is used to move the ferromagnetic particles toward the magnetic trapping component 8.4. Let the particle volume be Vp, the magnetic susceptibility difference between the particle and the fertilizer solution be Δχ, the vacuum permeability be μ0, and the magnetic induction intensity near the magnetic trapping component 8.4 be B. Then, by calculating the gradient of the magnetic field energy with respect to spatial position, the magnetic force Fm experienced by the particle is approximately: ; In the formula, Fm is the magnetic force on the magnetic particles, Vp is the volume of the magnetic particles, Δχ is the difference in magnetic susceptibility between the magnetic particles and the mixed fertilizer solution, and B is the magnetic induction intensity near the magnetic trapping component 8.4. The gradient operator is used to calculate the gradient with respect to spatial location. (B²) represents the gradient of the square of the magnetic induction intensity along the spatial position, and μ0 is the permeability of free space. Its subscript 0 indicates the vacuum state, so as to distinguish it from μ which represents the dynamic viscosity.

[0054] The magnetic field itself is not considered an independent fluid pressure term. The pressure drop in the second-stage magnetic particle separation device 8 mainly comes from the channel contraction and expansion, turning, magnetic trapping component 8.4, and the flow resistance formed by the already trapped particles. Therefore, the following approach is still used: ; For the third-stage automatic backwashing disc filter assembly 11, the multi-layered grooved discs form parallel tortuous microchannels, and their viscous friction, shrinkage, and turning losses are attributed to the same equivalent model: ; Substituting Q1=Q2=Q3=Q under normal series filtration conditions, and adding the pressure drops of the three stages one by one, we obtain the total pressure-flow relationship of the three-stage filtration system: ; Let aΣ = a1 + a2 + a3 and bΣ = b1 + b2 + b3, meaning aΣ is the sum of the equivalent coefficients of the three levels of viscous resistance and bΣ is the sum of the equivalent coefficients of the three levels of inertial resistance. Then, the normal filtration flow rate can be calculated by inversely using the inlet and outlet pressures: ; When bΣ is negligible, the above equation degenerates into: ; If the system only retains three outlet pressure measuring points, P1, P2, and P3, then only P1−P2, P2−P3, and P1−P3 can be directly obtained, and P0−P1 of the first-stage screen filter 7 cannot be independently obtained. Therefore, in this embodiment, the existing filter column inlet pressure sensor 10 is arranged before the inlet of the first-stage screen filter 7 and serves as the P0 measuring point. The above derivation shows the source relationship between pressure measuring points, flow measuring points, filter medium state, and clogging residual.

[0055] To reduce the possibility of viscosity changes being misinterpreted as filter clogging, controller 19 establishes a cleaning baseline differential pressure model for each stage of the filter assembly: ; And calculate the pressure-flow residual: ; Where i is the filter stage number, i=1, 2, 3; ΔPi is the measured pressure difference between the inlet and outlet of the i-th filter component; Quotient is the baseline pressure difference of the i-th stage filter element in a clean state. The "^" above the symbol indicates the predicted value calculated by the model. ai(μ, T) represents the viscous resistance equivalent coefficient as a function of dynamic viscosity μ and temperature T. ei is the pressure-flow residual of the i-th stage filter. ai and bi are obtained through equipment clean state calibration, historical operating data, or computational fluid dynamics analysis. If e1 continues to rise, it is first determined that the load on the first-stage screen filter 7 has increased. If e2 continues to rise, it is first determined that the capture load or flow resistance of the second-stage magnetic particle separation device 8 has increased. If e3 continues to rise and Q3 decreases, it is first determined that colloids, agglomerates, or filter cakes have formed on the surface of the stacked filter element 11.3.

[0056] The blockage diagnosis module can employ sliding window mean, rate of change, cumulative sum control chart, threshold combination, or machine learning classification model. If the pressure changes abruptly but the flow rate does not change synchronously, the controller 19 prioritizes judging sensor abnormalities, bubble or valve status abnormalities; if the pressure differentials at each level are within the normal range but the total output flow rate is inconsistent with the sum of the flow rates of each field fertilization branch, then the output manifold 22.1, output electric proportional valve 22.2-i, branch flow sensor 22.3-i, and field fertilization branch 22.4-i are checked first.

[0057] The backwash supply assembly 20 may include a backwash pump or pressure accumulator 20.2, a backwash check valve 20.3, and a filter isolation valve 20.4. The backwash medium is preferably from the clean water tank 3 or a clean water source that meets filtration requirements. The reverse pulse solenoid valve 13 generates a reverse pulse using a rapid opening and closing method; the controller 19 can adjust the pulse frequency, pulse duration, and number of repetitions according to the pressure difference and flow recovery degree before and after flushing, using proportional-integral-derivative control or segmented control.

[0058] During flushing, controller 19 first reduces or stops the adjustable speed hose pump 6.1, closes the recirculation valve 18, the normal output branch, and the isolation valve conflicting with the flushing direction; then it opens the drain valve 15 and opens the reverse pulse solenoid valve 13 according to the set sequence, allowing the flushing medium to pass through the stacked filter element 11.3 in the reverse direction from the clear liquid chamber 11.4. The stripped humic acid colloids, bacterial agent clusters, precipitates, or crystals enter the sludge collection area 16 and are discharged through the drain pipe 14. After flushing is completed, controller 19 closes the reverse pulse solenoid valve 13 and the drain valve 15, and re-collects P2, P3, and Q3; if the recovery conditions are met, normal filtration is restored; otherwise, flushing is repeated or an alarm is triggered and the system is shut down.

[0059] like Figure 6 As shown, the controller 19 can acquire one or more of the following data via wired or wireless interfaces: UAV multispectral remote sensing data, soil sensor network data, meteorological monitoring data, crop growth data, and knowledge of the crop's water and fertilizer requirements throughout its entire growth period. The intelligent decision-making module generates irrigation volume, fertilizer solution ratio, mixing concentration, application rate, and application timing based on the above data, and controls the multi-source metering and application component 21 and the multi-channel output control component 22.

[0060] The multi-channel output control component 22 preferably has 4 to 8 field fertilization branches, with each branch numbered n. The controller 19 adjusts the opening of the corresponding output electric proportional valve 22.2-i based on feedback from the flow sensors 22.3-i of each branch, ensuring that different fields or irrigation units receive independent target flow rates. During startup, refilling, and sampling stabilization phases, the controller 19 prioritizes opening the recirculation valve 18 and closing the field output; once the fertilizer solution is in suitable condition, it then closes the recirculation valve 18 and gradually opens each field fertilization branch.

[0061] The original liquid tank 2 can be equipped with a light-shielding outer shell, a heat-insulating jacket, and a phase change cold storage component 24. Assuming the fertilizer solution inside the tank has approximately uniform temperature after low-shear stirring, V is the effective volume of the fertilizer solution in the original liquid tank 2, and ρ is the density of the mixed fertilizer solution, then the mass mf of the fertilizer solution is obtained from the density and the effective volume: ; When ρ and cp remain approximately constant within the target temperature range, the sensible heat change dEf corresponding to the temperature change dT of the fertilizer solution is: ; In the formula, mf is the mass of fertilizer solution in the tank, V is the effective volume of fertilizer solution in the tank, cp is the specific heat capacity at constant pressure of the mixed fertilizer solution, dT is the small change in temperature of the fertilizer solution, and dEf is the change in sensible heat of the fertilizer solution corresponding to the temperature change.

[0062] Within a tiny time interval dt, the fertilizer solution gains heat from environmental heat transfer, solar radiation, and the operation of the agitator and pump, while the phase change cold storage module absorbs heat from the fertilizer solution. Let t be time, dt be the tiny time interval, and dEpcm / dt be the instantaneous heat absorption power of the phase change cold storage module. According to the first law of thermodynamics, we obtain: ; In the formula, the first term comes from Newton's law of cooling, representing the heat transfer rate between the environment and the fertilizer solution inside the tank. U is the total heat transfer coefficient of the tank wall of the raw liquid tank 2, together with the shading shell and the insulation layer. Ah is the effective heat exchange area of ​​the tank body, and Ta is the ambient temperature. The second term is the solar radiation input after the shading structure is corrected. γ is the effective absorption coefficient of the fertilizer solution to solar radiation after the shading structure is corrected. I is the solar irradiance, and Arad is the effective projected area of ​​the raw liquid tank 2 receiving solar radiation. The third term is the internal heat input of the agitator, pump, and other components that is ultimately converted into heat in the fertilizer solution. ηw is the heat conversion factor of this internal heat input, and Waux is the operating power of the agitator 4, pump, and other auxiliary components. To conservatively ensure that the fertilizer solution temperature does not exceed the target upper limit Tmax, only the positive heat load that raises the temperature of the fertilizer solution is considered, and it is integrated over the operating period from 0 to τ, where τ is the operating period duration within one configuration cycle. If the initial liquid temperature T0 is higher than the target upper limit Tmax, the initial excess sensible heat E0 must be removed first. ; Therefore, the conservation expression for the required cold storage energy Ereq during the configuration cycle is: ; In the formula, E0 is the initial superheated sensible heat that needs to be removed when the initial temperature of the fertilizer solution is higher than the target upper limit, T0 is the initial temperature of the fertilizer solution, and Tmax is the target upper limit; Ereq is the cold storage energy required in one configuration cycle; Ks is the cold storage safety factor, which is not less than 1 and is used to cover environmental fluctuations and parameter errors; ∫0τ(·)dt represents the time integral over the operation period from 0 to τ; Ta(t), I(t) and Waux(t) are the values ​​of ambient temperature, solar irradiance and auxiliary component operating power as a function of time, respectively.

[0063] When Ta, I, and Waux can be treated as constants within a configuration cycle, the terms are extracted from the integral sign to obtain the engineering configuration formula: ; The terms in this formula correspond to the initial overheat load, the environmental heat transfer load, the solar radiation load, and the internal heat generation load, respectively.

[0064] The usable cold storage capacity per unit mass of a phase change material (PCM) within the target temperature range consists of the sensible heat before the phase change, the latent heat of the phase change, and the sensible heat after the phase change. Let the phase change temperature be Tm, the latent heat of the phase change (i.e., the enthalpy difference of the phase change) be ΔHm, the specific heat capacities before and after the phase change be cs and cl, respectively, and the initial temperature of the PCM be Tpcm,0. Then the usable cold storage capacity per unit mass, ueff, is: ; In the formula, ueff is the available cold storage capacity per unit mass of the phase change material within the target temperature range, ΔHm is the latent heat of phase change, Tm is the phase change temperature, cs and cl are the solid-phase heat capacity before phase change and the liquid-phase heat capacity after phase change, respectively, and Tpcm,0 is the initial temperature of the phase change material. Considering the cold storage utilization efficiency ηpcm caused by module installation location, contact thermal resistance, and uneven heat transfer, the available cold storage capacity Epcm,usable that a phase change material with a mass of mpcm can provide is: ; In the formula, ηpcm is the cold storage utilization efficiency of the phase change cold storage module, mpcm is the configuration mass of the phase change material, and Epcm,usable is the available cold storage capacity that the phase change material can provide. Assuming the available cold storage capacity is not less than Ereq, the configuration mass of the phase change material mpcm and the number of phase change cold storage modules N are solved respectively: ; Where N is the number of phase change energy storage modules configured, and the symbol ⌈·⌉ indicates rounding up; mmod represents the mass of phase change material in a single phase change energy storage module. When the phase change temperature is not in the target temperature range, or when the material is undercooled, phase-separated, or has encapsulation heat transfer limitations, ueff does not directly use the nominal latent heat of the material, but is determined by the measured enthalpy difference or calibration curve.

[0065] When the phase change heat storage module has not yet absorbed heat and the environmental conditions are approximately constant, neglecting the phase change heat absorption term, the first-order temperature rise equation can be obtained from the above energy balance. Define the temperature response time constant τT and the equilibrium temperature without heat storage T∞: ; Where τT is the temperature response time constant of the fertilizer solution in the tank, T∞ is the equilibrium temperature of the fertilizer solution in the tank without the phase change cold storage module; t is the time elapsed after the start of the operation, and T(t) is the temperature of the fertilizer solution at time t; its temperature analytical relationship is: ; In the formula, exp(·) represents an exponential function with the natural logarithm as the base. From this analytical relationship, it can be seen that when T∞ is higher than Tmax, the fertilizer solution without a cold storage module will gradually approach an equilibrium temperature higher than the allowable upper limit. Therefore, a phase change cold storage module needs to be configured according to Ereq. When T∞ is not higher than Tmax, the controller can reduce the number of modules, but a cold storage safety factor Ks should still be retained to cover environmental fluctuations and parameter errors.

[0066] Different fertilizer solutions exhibit variations in viscosity, particle size distribution, particle settling velocity, and flocculation characteristics. Computational fluid dynamics can be used to analyze different fertilizer solution combinations, flow velocities, pipe diameters, bends, filter channels, and backwashing paths, and the simulation results can be applied to the design of sensor locations, channel structures, and control parameters. AI, Bi, clogging trigger thresholds, pulse frequency, flushing duration, and recovery thresholds should be determined through laboratory calibration, field trials, or historical operational data.

[0067] The integrated water and fertilizer platform can be used for field application of crops such as wheat and corn. Its application accuracy, anti-clogging performance, operational stability, online monitoring reliability, and operational efficiency can be evaluated at three levels: laboratory testing, field trials, and demonstration applications.

[0068] like Figure 7 and Figure 8 As shown, a control method for a mobile fertigation platform adapted to green fertilizers includes the following steps: S1 moves the integrated water and fertilizer platform to the operation point, connects to the field fertilization branch, and inputs the target irrigation volume, target ratio, target concentration, target pH range, target application time, and various detection and control parameters; and obtains one or more of the following information: crop growth, soil condition, weather condition, and crop growth period information, and generates a dynamic application prescription for water, humic acid stock solution, liquid microbial fertilizer, conventional fertilizer, and soluble carbon source. S2 controls the original liquid component and the multi-source metering and application component to send the corresponding materials into the mixing tank for mixing to obtain fertilizer solution; S3, through the feedback detection component, collects the fertilizer solution viscosity μ or equivalent viscosity, temperature T, and pH value, and simultaneously collects the conductivity EC. The collected data is filtered, temperature compensated, and outlier removed. The current fertilizer solution concentration K is calculated based on the temperature-compensated viscosity μ (i.e., μT), pH value, temperature T, and conductivity EC, using a regression model, lookup table model, interpolation model, or a calibrated machine learning model. =f(μ, pH, T, EC); S4, determine whether the current fertilizer solution concentration and pH value are within the corresponding target range; if not, turn off the multi-channel output control component and turn on the circulation return valve to allow the mixed fertilizer solution to return to the proportioning mixing tank through the circulation return pipe. Adjust one or more of the following: the metering flow rate of the original solution component and the multi-source metering and application component, the stirring frequency of the proportioning mixing tank, and the mixing time. After the delay stabilizes, return to step S3. S5, when the current fertilizer solution concentration and pH value are within the corresponding target range, close the circulation return valve, control the quantitative delivery component to make the fertilizer solution pass through the first-stage screen filter, the second-stage magnetic particle separation device and the third-stage automatic backwashing disc filter component in sequence, and then be delivered to the field fertilization branch through the multi-channel output control component. S6 collects the filtration pressure and flow rate of each stage of the three-stage filtration assembly, and determines the risk of clogging based on the coupling relationship between pressure and flow rate; Specifically, the inlet pressure of the front-end filter column of the three-stage filtration assembly is denoted as P0, and the outlet pressures of the first-stage screen filter, the second-stage magnetic particle separator, and the third-stage automatic backwashing disc filter assembly are denoted as P1, P2, and P3 respectively, with the corresponding outlet flow rates denoted as Q1, Q2, and Q3 respectively; the pressure differences between each stage are ΔP1=P0-P1, ΔP2=P1-P2, and ΔP3=P2-P3 respectively; the controller is based on... Determine the clean baseline pressure differential for the i-th stage of filtration, and based on... Determine the pressure-flow residual, where i is the filter stage number, and ai and bi are coefficients obtained from equipment calibration, historical operating data, or computational fluid dynamics analysis; One or more of the following methods can be used to identify congestion trends: sliding window mean, rate of change, cumulative sum control chart, threshold combination, or machine learning classification model; S7, when the blockage risk meets the preset triggering conditions, reduce or stop the quantitative delivery component, close the circulation return valve and the multi-channel output control component, open the reverse pulse solenoid valve, the drain valve and the backwash liquid supply component, so that the flushing medium passes through the clear liquid chamber of the third-stage automatic backwash disc filter component in the reverse direction and discharges the pollutants through the drain pipe. S8: After flushing, re-collect the pressure and flow of each stage of filtration. When the pressure and flow return to the preset range, resume normal output of step S5; otherwise, repeat flushing, alarm, or shutdown. Record the application prescription, sensor data, valve and pump control status, blockage warning, and flushing results for subsequent parameter calibration and operation strategy updates.

[0069] Furthermore, when normal filtration is performed and the recirculation valve, reverse pulse solenoid valve, and drain valve are all closed, Q1≈Q2≈Q3 is used as the flow continuity verification condition; when the recirculation valve, reverse pulse solenoid valve, drain valve, or multi-channel output control component is open, the corresponding recirculation flow, flushing flow, drain flow, or branch flow is added to the mass conservation relationship, and the diagnostic model is switched to the corresponding operating state.

Claims

1. A mobile fertigation platform adapted for green fertilizers, comprising a mobile support component (1), wherein the mobile support component (1) is equipped with a controller (19) for centralized control of the fertigation platform, characterized in that, The mobile carrier component (1) is connected in sequence to the proportioning and mixing component, the quantitative conveying component (6), and the three-stage filtration component via pipelines along the material conveying direction. The three-stage filtration component is equipped with pressure detection devices and flow detection devices to obtain the filtration pressure and flow rate of each stage. The proportioning and mixing assembly includes a proportioning and mixing tank (5), which is equipped with a stirrer (4). The inlet end of the proportioning and mixing tank (5) is connected to the original liquid assembly and the multi-source metering and dispensing assembly (21) through a manifold (5.1). A feedback detection component (9) is provided on the connecting pipe between the proportioning and mixing component and the quantitative delivery component (6). The feedback detection component (9) includes a viscosity detection unit (9.1), a temperature sensor (9.2), a pH sensor (9.3), and a conductivity sensor (9.4). The quantitative conveying component (6) is connected to the three-stage filtration component by a circulation return pipe (17) that is connected to the return port of the proportioning mixing tank (5). The circulation return pipe (17) is equipped with a circulation return valve (18). The three-stage filtration assembly includes a first-stage screen filter (7), a second-stage magnetic particle separator (8), and a third-stage automatic backwashing disc filter assembly (11) connected sequentially along the material conveying direction. The third-stage automatic backwashing disc filter assembly (11) includes a filter housing (11.1). The upper part of the filter housing (11.1) is formed by the disc filter element (11.3) to form an inlet chamber (11.2) and a clear liquid chamber (11.4). The lower part of the filter housing (11.1) is provided with a dirt collection area (16) that communicates with the inlet chamber (11.2). The outlet end of the dirt collection area (16) is connected to a drain pipe (14), and a drain valve (15) is provided on the drain pipe (14). The clear liquid chamber (11.4) is connected to a multi-channel output control assembly (22) through an output pipe (12) and to a backwashing supply assembly (20) through a backwashing pipe (20.1). A reverse pulse solenoid valve (13) is provided on the backwashing pipe (20.1).

2. The mobile fertigation platform adapted for green fertilizers according to claim 1, characterized in that, The raw liquid assembly includes a raw liquid tank (2) and a clean water tank (3). The outlet end of the raw liquid tank (2) is connected to the manifold (5.1) through a raw liquid outlet pipe (2.1), and the outlet end of the clean water tank (3) is connected to the manifold (5.1) through a clean water outlet pipe (3.1). The raw liquid tank (2) is equipped with a light-shielding and heat-insulating phase change cold storage assembly (24). The light-shielding and heat-insulating phase change cold storage assembly (24) includes at least one of a light-shielding shell, a heat-insulating interlayer, and a replaceable phase change cold storage module. The multi-source metering and application assembly (21) includes several multi-source application branches (21.1-i). Each multi-source application branch (21.1-i) is sequentially connected to a branch metering pump (21.2-i), an application electric proportional valve (21.3-i), a branch flow sensor (21.4-i), and a branch check valve (21.5-i). All multi-source application branches (21.1-i) are connected to the manifold (5.1) through a common application manifold (21.6).

3. The mobile fertigation platform adapted for green fertilizers according to claim 1, characterized in that, The pressure detection device includes a filter column inlet pressure sensor (10) located before the inlet of the first-stage screen filter (7), and a first pressure sensor (7.1), a second pressure sensor (8.1), and a third pressure sensor (11.5) located at the outlet of the first-stage screen filter (7), the outlet of the second-stage magnetic particle separator (8), and the outlet of the third-stage automatic backwashing disc filter assembly (11), respectively; the flow detection device includes a first flow sensor (7.2), a second flow sensor (8.2), and a third flow sensor (11.6) located at the outlet of the first-stage screen filter (7), the outlet of the second-stage magnetic particle separator (8), and the outlet of the third-stage automatic backwashing disc filter assembly (11), respectively.

4. The mobile fertigation platform adapted for green fertilizers according to claim 1, characterized in that, The multi-channel output control component (22) includes an output manifold (22.1) and several field fertilization branches (22.4-i). The output manifold (22.1) is connected to the output pipe (12). All field fertilization branches (22.4-i) are connected in parallel to the output manifold (22.1). Each field fertilization branch (22.4-i) is equipped with an output electric proportional valve (22.2-i) and a branch flow sensor (22.3-i).

5. A mobile fertigation platform adapted for green fertilizers according to claim 1, characterized in that, The backwash supply assembly (20) includes a filter isolation valve (20.4), a backwash check valve (20.3), and a backwash pump (20.2) connected in sequence in the direction away from the reverse pulse solenoid valve (13).

6. A control method for a mobile fertigation platform adapted to green fertilizers as described in claim 2, characterized in that, Includes the following steps: S1, move the integrated water and fertilizer platform to the work site, connect the multi-channel output control component (22), and initialize the target ratio of fertilizer solution, target concentration, target pH range, allowable error and sampling cycle; S2, control the original liquid component and the multi-source metering and application component (21) to send the corresponding materials into the mixing tank (5) for mixing to obtain fertilizer solution; S3, collect the fertilizer solution viscosity, temperature and pH value through the feedback detection component (9), filter and temperature compensate the collected data, and estimate the current fertilizer solution concentration; S4, determine whether the current fertilizer solution concentration and pH value are within the corresponding target range; if not within the corresponding target range, close the multi-channel output control component (22), open the circulation reflux valve (18), adjust one or more of the following: the metering flow rate of the original solution component, the multi-source metering and application component (21), the mixing frequency and mixing time of the mixing tank (5), and return to step S3 after a delay and stabilization. S5, when the current fertilizer concentration and pH value are within the corresponding target range, close the circulation reflux valve (18), and control the quantitative delivery component (6) to make the fertilizer solution pass through the first-stage screen filter (7), the second-stage magnetic particle separation device (8) and the third-stage automatic backwashing disc filter component (11) in sequence before being output; S6 collects the filtration pressure and flow rate of each stage of the three-stage filtration assembly, and determines the risk of clogging based on the coupling relationship between pressure and flow rate; S7, when the blockage risk meets the preset triggering conditions, reduce or stop the quantitative delivery component (6), close the circulation return valve (18) and the multi-channel output control component (22), open the reverse pulse solenoid valve (13), the drain valve (15) and the backwash liquid supply component (20), so that the flushing medium passes through the clear liquid chamber (11.4) of the third-stage automatic backwash disc filter component (11) in the reverse direction through the disc filter element (11.3) and discharges the pollutants through the drain pipe (14); S8: After rinsing, re-collect the pressure and flow of each stage of the filter. When the pressure and flow return to the preset range, resume normal output of step S5; otherwise, repeat rinsing, alarm, or shutdown.

7. The control method for a mobile fertigation platform adapted to green fertilizers according to claim 6, characterized in that, S3 also includes the collection of electrical conductivity EC and current fertilizer solution concentration. Based on the temperature-compensated viscosity μ, pH value, temperature T, and conductivity EC, the following parameters are calculated using regression models, lookup table models, interpolation models, or calibrated machine learning models: =f(μ,pH,T,EC).

8. The control method for a mobile fertigation platform adapted to green fertilizers according to claim 6, characterized in that, In S6, the inlet pressure of the front filter column of the three-stage filter assembly is recorded as P0, and the outlet pressures of the first-stage screen filter (7), the second-stage magnetic particle separator (8), and the third-stage automatic backwashing disc filter assembly (11) are recorded as P1, P2, and P3 respectively, and the corresponding outlet flow rates are recorded as Q1, Q2, and Q3 respectively; the pressure differences of each stage are ΔP1=P0-P1, ΔP2=P1-P2, and ΔP3=P2-P3 respectively; the controller (19) according to Determine the clean baseline pressure differential for the i-th stage of filtration, and based on... Determine the pressure-flow residual, where ai and bi are coefficients obtained from equipment calibration, historical operating data, or computational fluid dynamics analysis.

9. The control method for a mobile fertigation platform adapted to green fertilizers according to claim 6, characterized in that, S6 uses one or more of the following methods to identify congestion trends: sliding window mean, rate of change, cumulative sum control chart, threshold combination, or machine learning classification model.

10. The control method for a mobile fertigation platform adapted to green fertilizers according to claim 8, characterized in that, When normal filtration is performed and the recirculation valve (18), the reverse pulse solenoid valve (13), and the drain valve (15) are all closed, Q1≈Q2≈Q3 is used as the flow continuity verification condition; when the recirculation valve (18), the reverse pulse solenoid valve (13), the drain valve (15), or the multi-channel output control component (22) is open, the corresponding recirculation flow, flushing flow, drain flow, or branch flow is added to the mass conservation relationship, and the diagnostic model of the corresponding operating state is switched.

Images