A Water and Fertilizer Balanced Supply Control Method Based on Photovoltaic-Storage Combined Energy Supply
By using a combined photovoltaic and energy storage system to control the balanced supply of water and fertilizer, and employing a multivariate coupling model and composite control technology, the instability and mixing uniformity of the water and fertilizer system under light fluctuations are solved. This achieves high stability and high uniformity of water and fertilizer concentration and flow, protects the energy storage system, and adapts to the agronomic needs of different operating conditions.
Patent Information
- Application Number
- CN202611439765.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-16
- Publication Date
- 2026-11-17
AI Technical Summary
Existing solar-powered integrated water and fertilizer systems suffer from unstable water and fertilizer output under fluctuating light conditions. They lack proactive compensation control with feedforward sensing of light changes, and there is a lack of coordinated optimization between water and fertilizer ratio and energy management. Mixing uniformity is difficult to guarantee, and the control method is passive and reactive, lacking the ability to predict and proactively compensate for future disturbances.
A water and fertilizer balance supply control method based on photovoltaic and energy storage is adopted. Through rolling optimization of a multivariate coupled model, energy distribution and water and fertilizer ratio are actively coordinated. Combined with composite control of feedforward compensation and feedback regulation, the mixer is adaptively adjusted to ensure mixing uniformity over a wide flow range.
It achieves high uniformity and stability in water and fertilizer concentration and flow under complex lighting conditions, reduces the impact of light fluctuations on water and fertilizer output, protects the energy storage system, adapts to dynamic characteristics under different operating conditions, and ensures the basic water and fertilizer needs of crops.
Smart Images

Figure CN123113569A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural intelligent equipment and automatic control technology, specifically to a water and fertilizer balanced supply control method based on photovoltaic and energy storage combined power supply. Background Technology
[0002] Integrated water and fertilizer management technology is an important agricultural measure to achieve water and fertilizer conservation, increased yield and efficiency. In agricultural production areas with insufficient power grid coverage, such as hilly and mountainous areas, solar photovoltaic power generation has become the main energy source for driving irrigation and fertilization equipment. However, existing solar-powered integrated water and fertilizer management systems generally suffer from the following prominent problems:
[0003] 1. Fluctuations in light intensity lead to unstable water and fertilizer output:
[0004] The output power of solar photovoltaic systems fluctuates dramatically with factors such as solar radiation intensity, weather, and cloud cover, directly causing pulsations in water pump speed and fertilizer pump flow rate, resulting in drastic variations in water-fertilizer mixing concentration and output flow rate. Existing systems mostly use batteries for direct power supply, which only serves to smooth out peak and valley loads, lacking proactive compensation control for water and fertilizer output based on feedforward sensing of changes in sunlight. Consequently, the uniformity of water and fertilizer distribution drops sharply during sudden changes in sunlight.
[0005] 2. Lack of synergistic optimization between water and fertilizer ratios and energy management:
[0006] Traditional fertigation systems primarily rely on feedback regulation using EC / pH sensors, performing closed-loop correction only for fertilizer concentration. However, under combined solar and energy storage power supply conditions, the available solar energy, the state of charge (SOC) of the storage battery, the pump flow rate, and the opening of the fertilization channel form a mutually constrained relationship. Single-concentration feedback regulation cannot reconcile the conflict between energy allocation and agronomical demands. For example, when sunlight suddenly drops or the battery SOC is low, continuing to fertilize at the original concentration and flow rate may lead to fertilization interruption due to insufficient power; conversely, simply reducing the flow rate will cause the fertilizer concentration to deviate from the target range.
[0007] 3. Lack of prediction-based feedforward-feedback composite control: Most existing systems are reactive control, that is, the fertilizer pump or valve is adjusted only after the concentration deviation is detected, which results in regulation lag.
[0008] 4. Mixing uniformity is difficult to guarantee under variable flow conditions:
[0009] Under varying flow and pressure conditions, the internal flow field of the static mixer in the water and fertilizer head system deteriorates, easily leading to fertilizer crystallization, stratification, or uneven mixing. Especially under frequent start-ups and shutdowns and sudden flow changes caused by fluctuations in photovoltaic and energy storage supply, traditional static mixers cannot adaptively adjust, severely affecting the uniformity of fertilizer solution. Summary of the Invention
[0010] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a water and fertilizer balanced supply control method based on photovoltaic and energy storage combined power supply. This method can: actively coordinate energy distribution and water and fertilizer ratio through rolling optimization of a multivariate coupled model based on photovoltaic forecasts and energy storage status; achieve high uniformity and high stability output of water and fertilizer concentration and flow rate under complex lighting conditions by combining feedforward compensation and feedback regulation composite control; automatically switch to a minimum protection mode under low energy conditions to ensure the minimum water and fertilizer requirements of crops while protecting the energy storage system; and adaptively adjust the mixer according to real-time flow rate changes to ensure mixing uniformity over a wide flow rate range.
[0011] To achieve the above-mentioned objectives, this invention provides a water and fertilizer balanced supply control method based on photovoltaic and energy storage combined power supply. This method is applied to a system including a photovoltaic array, an energy storage unit, a water and fertilizer head unit, a sensor group, and a controller. The photovoltaic array and the energy storage unit jointly power the water and fertilizer head unit via DC bus coupling. The controller is electrically connected to both the sensor group and the water and fertilizer head unit. The water and fertilizer head unit includes at least a variable frequency water pump, an adjustable fertilizer channel, a mother liquor container, and a static mixer. The adjustable fertilizer channel is connected to the mother liquor container and the water supply pipeline. The static mixer is located between the outlet of the variable frequency water pump and the junction of the adjustable fertilizer channel, and is used to mix the mother liquor with the water flow to output fertilizer solution; the sensor group includes at least: a photovoltaic irradiance sensor for detecting light intensity, an energy storage state of charge detection module for detecting the remaining power of the energy storage unit, a conductivity / pH sensor for detecting the concentration of the mixed fertilizer solution, and an outlet flow meter for detecting the system output flow rate; the controller receives externally input irradiance prediction data; the method includes the following steps:
[0012] Step S1: Obtain the current state of charge of the energy storage unit through the energy storage state of charge detection module, and receive the irradiance prediction data for future time periods through the controller to determine the photovoltaic power prediction value;
[0013] Step S2: Generate a reference trajectory for the target fertilizer solution concentration and target irrigation flow rate based on the crop's water and fertilizer requirements;
[0014] Step S3: Based on the pre-established multivariate dynamic model, the photovoltaic power prediction value is used as the input of the multivariate dynamic model. The current state of charge, the current concentration of the mixed fertilizer solution detected by the conductivity / alkalinity sensor, and the current total flow rate detected by the outlet flow meter are used as state variables. Under the condition of satisfying the energy storage state of charge constraint and power balance constraint, the reference trajectory is used as the optimization objective to perform rolling optimization solution to obtain the optimal control sequence of water pump frequency and fertilizer channel opening.
[0015] Step S4: During the execution of the control sequence to drive the variable frequency water pump and the adjustable fertilizer channel, the actual photovoltaic power is detected in real time by the photovoltaic irradiance sensor. When the fluctuation of the actual photovoltaic power exceeds the preset threshold, the feedforward compensation amount is determined according to the preset mapping relationship with the photovoltaic power mutation amount as input and the water pump frequency compensation amount and the fertilizer channel opening compensation amount as output. The feedforward compensation amount is superimposed on the control sequence, and the variable frequency water pump and the adjustable fertilizer channel are continued to be driven by the superimposed control sequence.
[0016] Step S5: The actual concentration of the mixed fertilizer solution is detected in real time by the conductivity / pH sensor. Based on the deviation between the actual concentration and the target fertilizer solution concentration in the reference trajectory, the opening of the fertilization channel is adjusted accordingly.
[0017] Existing integrated photovoltaic-storage power systems with combined water and fertilizer management suffer from three core problems: first, water and fertilizer output fluctuates drastically with sunlight, resulting in poor stability of concentration and flow rate; second, the control method is passive and reactive, lacking the ability to predict and actively compensate for future disturbances; and third, energy management and water and fertilizer ratio are independently regulated, lacking coordinated optimization. This invention addresses these issues by: Step S1 acquiring current state and predicted data through sensors to provide a foundation for subsequent optimization; Step S2 generating a reference trajectory to provide an optimization target; Step S3 performing rolling optimization based on a multivariate dynamic model, using the predicted photovoltaic power as model input and the reference trajectory as the optimization target, to solve for the optimal control sequence under the constraints of energy storage and power balance; Step S4 rapidly adding compensation amounts during power surges through feedforward mapping; and Step S5 performing adaptive PID control through concentration feedback. These five steps form a complete prediction-optimization-execution-compensation-correction closed loop. Through this five-step control method, highly stable output of water and fertilizer concentration and flow rate is achieved. Specifically, this manifests as follows: proactive optimization through prediction significantly reduces the impact of light fluctuations on water and fertilizer output; feedforward-feedback composite control balances the speed of disturbance suppression with the accuracy of steady-state correction; and multivariate coupling models enable coordinated optimization of energy scheduling and water-fertilizer ratio within a unified framework.
[0018] Preferably, the multivariate dynamic model is composed of the following coupled sub-models:
[0019] The power balance sub-model is used to describe the dynamic balance relationship between photovoltaic output power, energy storage charging and discharging power, and water pump power consumption.
[0020] P pv +P bat =P pump +P fert +P loss Among them, P pv P is the output power of the photovoltaic array.bat For the charging and discharging power of the energy storage unit, P pump For the power consumption of the variable frequency water pump, P loss For the system power loss, P fert Power consumption for adjustable fertilization channels;
[0021] The variable frequency pump flow-frequency-power sub-model describes the output flow rate of the variable frequency pump after its frequency is converted through nonlinear efficiency. The mapping relationship between this output flow rate and the pump head is then used to determine the power consumption of the variable frequency pump.
[0022] Q water =η1(f)×f,P pump =η2(f)×Q water ×H; where Q water η1(f) is the output flow rate of the variable frequency pump, f is the frequency of the variable frequency pump, η2(f) is the nonlinear efficiency function of flow rate-frequency, and H is the head of the variable frequency pump.
[0023] The dynamic sub-model for water and fertilizer mixing takes the mother liquor injection flow rate q(t) at time t and the variable frequency pump output flow rate Q at time t as inputs. water (t), the output is the actual fertilizer solution concentration C at the outlet of the static mixer at time t. out (t), the expression for the dynamic sub-model of water and fertilizer mixing is:
[0024] C(t) = q(t) × C m / (Q water (t)+q(t));
[0025] ;
[0026] Where C(t) is the theoretical concentration of the fertilizer solution after mixing at time t, C m Where is the concentration of the mother liquor, T is the mixing time constant, and τ is the lag time. For a moment The theoretical fertilizer solution concentration at that time;
[0027] P in the power balance sub-model pump and P fert The values of Q are determined by the control variables in the variable frequency pump flow-frequency-power sub-model and the water-fertilizer mixing dynamic sub-model, respectively; the output Q of the variable frequency pump flow-frequency-power sub-model is... water This serves as the input variable for the dynamic sub-model of water and fertilizer mixing.
[0028] The multivariate dynamic model is composed of a power balance sub-model, a variable frequency pump flow-frequency-power sub-model, and a water-fertilizer mixing dynamic sub-model. The power balance sub-model describes the energy distribution relationship, the variable frequency pump sub-model describes the frequency-to-flow conversion and power consumption mapping, and the water-fertilizer mixing dynamic sub-model describes the dynamic process of mixing the mother liquor with the water flow to form fertilizer concentration. The three sub-models are coupled through shared variables: P in the power balance sub-model... pump and P fert The Q output of the variable frequency pump sub-model is determined by the control variables in the variable frequency pump sub-model and the water-fertilizer mixing sub-model. water As input variables to the water-fertilizer mixture sub-model, the output C of the water-fertilizer mixture sub-model is... out This is the final output.
[0029] Preferably, when the energy storage unit discharges, P bat When the energy storage unit is charging, P is positive. bat If negative, P loss Constrained by the upper and lower limits of the energy storage state of charge (SOC); in the step of performing rolling optimization, the optimization objective function includes a penalty term for deviation of the energy storage state of charge from the preset safe range, and the constraints include the upper and lower limits of the SOC, the maximum discharge power limit of the energy storage unit, the frequency range of the variable frequency water pump, the opening range of the adjustable fertilizer channel, the upper and lower limits of the system output flow detected by the outlet flow meter, the upper and lower limits of the outlet pressure, and the allowable fluctuation range of the concentration.
[0030] This method incorporates a penalty term into the objective function of the rolling optimization solution for deviations of the energy storage state of charge (SOC) from a preset safe range, causing the algorithm to automatically tend to reduce energy consumption when the SOC is low. The constraints are comprehensive, including upper and lower limits of SOC, maximum discharge power limit, frequency range of the variable frequency pump, adjustable fertilizer channel opening range, upper and lower limits of system output flow rate, upper and lower limits of outlet pressure, and allowable concentration fluctuation range, thus fully defining the feasible region of the optimization problem. By adding a penalty term for deviations of the energy storage SOC from the preset safe range, the optimization solution aims to achieve both accurate water and fertilizer output and energy storage lifespan protection. The complete enumeration of constraints clarifies the boundaries of the feasible region of the optimization solution, ensuring the executability of the solution in a real system.
[0031] Preferred,
[0032] In step S2, when it is predicted that the predicted photovoltaic power value will continue to be lower than the total power consumption of the variable frequency water pump and the adjustable fertilizer channel in the future period, and the current state of charge drops to the lower limit threshold SOC, minWhen the target irrigation flow rate in the reference trajectory is switched from the current value to a preset minimum flow rate value, and the target fertilizer concentration is locked unchanged, the reference trajectory in the low-power minimum mode is generated; when the predicted photovoltaic power recovers and the current state of charge rises to the recovery threshold SOC... rec When this happens, the low-power backup mode is exited, and the normal reference trajectory is generated again.
[0033] Under normal operating conditions, this method can ensure a balanced supply of water and fertilizer. However, under extreme conditions, such as continuous rainy days leading to insufficient photovoltaic power generation and near-depletion of energy storage charge, adhering to the original irrigation flow and concentration targets may result in over-discharge and damage to the energy storage or system shutdown, posing a risk of crop water and fertilizer shortages. To address this issue, this method, during the rolling optimization process, considers situations where the predicted photovoltaic power is consistently lower than the total power consumption and the current state of charge (SOC) has fallen to the lower limit threshold. min At this time, the target irrigation flow rate in the reference trajectory is switched to the preset minimum flow rate value, and the target fertilizer concentration is locked unchanged to enter the low power consumption minimum mode. When the predicted photovoltaic power recovers and the state of charge rises to the recovery threshold SOC... rec When energy is severely insufficient, the system automatically exits the safety mode and resumes normal irrigation. By introducing a low-power safety mode, the irrigation flow rate is automatically reduced and the target concentration is locked when energy is severely insufficient, ensuring the basic water and fertilizer needs of crops with minimal energy consumption. When energy is restored, the system automatically exits the safety mode and resumes normal irrigation. This mechanism protects the energy storage system from damage due to over-discharge and ensures the survival needs of crops under extreme conditions.
[0034] Preferably, during the execution of steps S4 and S5, the method further includes: setting adjustable guide vanes before the static mixer, and adjusting the flow rate based on the real-time total flow rate Q detected by the outlet flow meter and the target fertilizer concentration C in the reference trajectory. ref Adjust the opening of the guide vanes to maintain the Reynolds number in the static mixer within a preset range.
[0035] Currently, the water-fertilizer mixing process relies on a static mixer. However, under photovoltaic energy storage conditions, the flow rate fluctuates with sunlight, causing the traditional static mixer to deteriorate in mixing efficiency under variable flow conditions, easily leading to fertilizer stratification or uneven mixing, affecting the uniformity of fertilizer concentration. To solve this problem, adjustable guide vanes are installed before the static mixer. During the execution of steps S4 and S5, the real-time total flow rate Q detected by the outlet flow meter and the target fertilizer concentration C in the reference trajectory are used as the basis for the mixing process. ref Adjust the opening of the guide vanes to maintain the Reynolds number Re in the static mixer within the preset optimal range [Re]. min Re maxBy adding adjustable guide vanes before the static mixer and adjusting the vane opening in real time according to flow rate and concentration, the Reynolds number within the mixer is always maintained within the optimal range, ensuring mixing uniformity over a wide flow range and solving the problem of decreased mixing quality under varying operating conditions. Specifically, Re... min Let Re be the lower limit Reynolds number of the optimal interval. max is the upper limit Reynolds number of the optimal interval.
[0036] Preferably, the mapping relationship is established through offline calibration experiments, with the input being the photovoltaic power mutation ΔP. pv The real-time total flow rate Q detected by the outlet flow meter is output as the pump frequency compensation amount Δf. ff Compensation amount Δu for fertilization channel opening ff The mapping relationship is stored in the memory of the controller.
[0037] Preferably, the rolling optimization solution employs a model predictive control algorithm, wherein the prediction time domain N p Covering at least one illumination fluctuation cycle, controlling the time domain N c Less than the prediction time domain N p .
[0038] The rolling optimization solution employs a model predictive control algorithm. Prediction time domain N p Covering at least one illumination fluctuation cycle allows the optimization to observe the complete illumination change trend; controlling the time domain N c The control time domain is shorter than the prediction time domain, ensuring timely control action without excessive anticipation. Optimization involves predicting future states based on photovoltaic power forecasts within the prediction time domain, and determining the optimal control sequence for pump frequency and fertilizer channel opening within the control time domain. The rolling optimization solution employs a model predictive control algorithm, with the prediction time domain covering the solar radiation fluctuation cycle and the control time domain shorter than the prediction time domain. This allows the rolling optimization solution to predict future states based on photovoltaic power forecasts within the prediction time domain and determine the optimal control sequence within the control time domain. It fully considers the periodic characteristics of solar radiation fluctuations, ensuring effective utilization of prediction information and timely control response.
[0039] Preferably, in step S2, the target fertilizer concentration and the target irrigation flow rate are dynamically adjusted based on the preset water and fertilizer plan for the current crop growth stage and the irradiance prediction data. Crops at different growth stages have different water and fertilizer requirements (e.g., seedlings require less water, while jointing requires more). Irradiance prediction data can reflect future evapotranspiration trends. Combining these two factors generates a reference trajectory that balances crop needs and energy constraints. By clarifying the basis for the dynamic adjustment of the reference trajectory, the generated reference trajectory can adaptively adjust according to the crop growth stage and light conditions, better meeting actual agronomic needs and improving the accuracy and adaptability of water and fertilizer management.
[0040] Preferably, in step S4, when a sudden change in photovoltaic power is detected, the water pump frequency compensation amount Δf ff and the fertilizer application channel opening compensation amount Δu ff The control sequence is superimposed before the concentration deviation occurs; in step S5, the feedback adjustment is achieved through an adaptive PID controller, and the parameters of the adaptive PID controller are adjusted online according to the rate of change of the real-time total flow rate Q detected by the outlet flow meter and the deviation between the actual concentration and the target fertilizer concentration in the reference trajectory.
[0041] In step S4, when a sudden change in photovoltaic power is detected, the water pump frequency compensation amount Δf ff Compensation amount Δu for fertilization channel opening ff Before the concentration deviation occurs, the feedforward intervention is achieved by directly superimposing the feedforward into the control sequence and utilizing the physical delay window between power change and concentration change. In step S5, feedback regulation is implemented through an adaptive PID controller. The PID parameters are adjusted online based on the real-time total flow rate Q detected by the outlet flow meter and the rate of change of concentration deviation. The control effect is enhanced when the flow rate is high and the deviation changes rapidly, and weakened when the flow rate is low, in order to adapt to different operating conditions. The feedforward compensation is directly superimposed into the control sequence before the concentration deviation occurs, thus clarifying the advance suppression characteristic of the feedforward. The online adjustment of the PID parameters based on the real-time total flow rate and the rate of change of concentration deviation enables the feedback regulation to adapt to the dynamic characteristics under different operating conditions, thereby improving the robustness of the feedback control.
[0042] Preferably, the sensor group includes an influent flow meter and a mother liquor flow meter; in step S3, the multivariate dynamic model uses the influent flow rate detected by the influent flow meter and the mother liquor injection flow rate detected by the mother liquor flow meter as inputs to update the state variable values in the multivariate dynamic model. Using the detected values to update the state variable values in the model ensures that the multivariate dynamic model is always optimized based on the latest actual measurement data, thus improving the accuracy and control precision of the model.
[0043] One or more technical solutions provided by this invention have at least the following technical effects or advantages:
[0044] 1. Highly stable water and fertilizer output:
[0045] By using photovoltaic forecasting feedforward and MPC to coordinate energy and materials in advance, the lag of simple feedback control is avoided, which greatly reduces the fluctuation of water and fertilizer concentration and significantly reduces flow fluctuation when there are severe light fluctuations.
[0046] 2. Multi-objective collaborative optimization:
[0047] An innovative energy-flow-fertilizer coupling model was established to achieve synergistic optimization of energy storage scheduling and water-fertilizer ratio. This ensures the precise amount of water and fertilizer required by crops while maximizing photovoltaic self-sufficiency, reducing the number of deep cycles of batteries, and extending the lifespan of the energy storage system.
[0048] 3. Significantly improved mixing uniformity:
[0049] By introducing an adaptive flow guide mixer and variable operating condition Reynolds number control, the uniformity of fertilizer solution mixing under different flow rates is always kept at an optimal level.
[0050] 4. Wide adaptability to complex scenarios:
[0051] The system is completely independent of the power grid, supports stable off-grid operation, and is robustly designed to address characteristics such as light fluctuations and dispersed plots, making it suitable for precision water and fertilizer management of dryland crops in hilly and mountainous areas.
[0052] 5. Low-energy consumption protection:
[0053] When sunlight is consistently insufficient and energy storage is nearly depleted, the irrigation flow rate is automatically reduced and the target concentration is locked to ensure the basic water and fertilizer needs of crops with minimal energy consumption, while preventing damage from excessive energy release. Attached Figure Description
[0054] The accompanying drawings, which are provided to further illustrate embodiments of the invention and constitute a part of this invention, are not intended to limit the scope of the invention.
[0055] Figure 1 This is a schematic diagram of a water and fertilizer balance supply control method based on photovoltaic and energy storage combined power supply in this invention. Detailed Implementation
[0056] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, where there is no conflict, the embodiments of the present invention and the features thereof can be combined with each other.
[0057] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.
[0058] Example 1;
[0059] Please refer to Figure 1 , Figure 1This invention provides a water and fertilizer balanced supply control method based on photovoltaic and energy storage combined power supply, applicable to a system including a photovoltaic array, an energy storage unit, a water and fertilizer head unit, a sensor group, and a controller. The photovoltaic array and the energy storage unit jointly power the water and fertilizer head unit via DC bus coupling. The controller is electrically connected to both the sensor group and the water and fertilizer head unit. The water and fertilizer head unit includes at least a variable frequency water pump, an adjustable fertilizer channel, a mother liquor container, and a static mixer. The adjustable fertilizer channel is connected to the sensor group and the water and fertilizer head unit. The mother liquor container is located between the water supply pipeline and the water supply pipeline, used to regulate the mother liquor injection flow rate; the static mixer is located after the junction of the outlet of the variable frequency water pump and the adjustable fertilizer channel, used to mix the mother liquor with the water flow and output fertilizer solution; the sensor group includes at least: a photovoltaic irradiance sensor for detecting light intensity, an energy storage state of charge detection module for detecting the remaining power of the energy storage unit, a conductivity / pH sensor for detecting the concentration of the mixed fertilizer solution, and an outlet flow meter for detecting the system output flow rate; the controller receives externally input irradiance prediction data; the method includes the following steps:
[0060] Step S1: Obtain the current state of charge of the energy storage unit through the energy storage state-of-charge detection module, and receive the irradiance prediction data for future periods through the controller to determine the photovoltaic power prediction value; the energy storage state-of-charge detection module is integrated into the battery management system of the battery pack. The battery management system uses a dedicated lithium battery management chip and communicates with the controller via an RS485 bus. The controller connects to the cloud platform through a communication module to obtain the irradiance prediction data for future periods from the cloud platform. The specific implementation of step S1 can be as follows: The controller sends a SOC read command to the battery management system via the RS485 bus; the battery management system returns the current SOC value; the controller performs CRC verification, and stores the current state of charge after the verification is successful; the controller checks whether the irradiance prediction data in the local memory is valid and covers the next 2 hours; if an update is needed, the controller requests the irradiance prediction data and ambient temperature prediction data for the next 2 hours from the cloud platform via the communication module; the controller converts the irradiance prediction data sequence point by point into a photovoltaic power prediction value sequence based on the photovoltaic array model parameters; the controller stores the current state of charge and photovoltaic power prediction value sequence in memory for use in step S3; wherein, the method for determining the photovoltaic power prediction value through the irradiance prediction data is: P pv (t)=P rated ×G pred (t) / G STC In the formula, P pv (t) represents the predicted photovoltaic power at time t, P rated For the installed capacity of photovoltaic arrays, G pred (t) represents the predicted irradiance at time t, GSTC This is the standard test irradiance.
[0061] Step S2: Generate a reference trajectory for the target fertilizer solution concentration and target irrigation flow rate based on the crop's water and fertilizer requirements;
[0062] The controller incorporates a crop growth calendar, automatically calculating the current growth stage based on the planting date set during system installation. Taking corn as an example, the growth stages are divided as follows: seedling stage (1-25 days after sowing), jointing stage (26-50 days after sowing), tasseling stage (51-70 days after sowing), grain-filling stage (71-100 days after sowing), and maturity stage (101-120 days after sowing). The controller obtains the current date through the system clock and compares it with the planting date to determine the crop's current growth stage.
[0063] The controller's memory pre-stores water and fertilizer schedules for different growth stages, which are based on local agronomic experiment data. In this embodiment, the pre-set water and fertilizer schedules for each growth stage of corn are as follows:
[0064] Seedling stage: Target fertilizer solution concentration EC=1.2mS / cm, single irrigation volume 8-12m³ / mu; Jointing stage: Target fertilizer solution concentration EC=2.0mS / cm, single irrigation volume 15-20m³ / mu; Tasseling stage: Target fertilizer solution concentration EC=2.5mS / cm, single irrigation volume 18-25m³ / mu; Grain filling stage: Target fertilizer solution concentration EC=2.0mS / cm, single irrigation volume 12-18m³ / mu; Maturity stage: Target fertilizer solution concentration EC=1.0mS / cm, single irrigation volume 8-10m³ / mu.
[0065] The controller reads the corresponding target fertilizer solution concentration baseline value C from the program table based on the current growth stage. base and the target irrigation flow baseline value Q base In this embodiment, the current stage is the jointing stage, so C base =2.0 mS / cm, Q base Take the median value of 17.5 m³ / mu.
[0066] Considering the impact of light conditions on crop evapotranspiration and water requirements, the controller further dynamically adjusts the baseline value of irrigation flow based on irradiance prediction data. In this embodiment, the controller calculates the average value of the predicted irradiance over the next two hours and compares it with the typical irradiance for the season.
[0067] If the average value is ≥1.3 × the typical irradiance of the season (strong sunlight), the target irrigation flow rate will be increased by 10%.
[0068] If the average value is ≤0.7 × typical irradiance for the season (weak sunlight), then the target irrigation flow rate will be reduced by 10%.
[0069] If 0.7 × typical irradiance for the season < average value < 1.3 × typical irradiance for the season (normal light), then the baseline value remains unchanged.
[0070] The target fertilizer solution concentration is not adjusted according to light conditions and remains at C. base The controller will adjust the target fertilizer solution concentration C obtained after the above adjustments. ref and target irrigation flow rate Q ref The reference trajectory for this control cycle is stored in memory for use in step S3.
[0071] Step S3: Based on the pre-established multivariate dynamic model, the photovoltaic power prediction value is used as the input of the multivariate dynamic model. The current state of charge, the current concentration of the mixed fertilizer solution detected by the conductivity / alkalinity sensor, and the current total flow rate detected by the outlet flow meter are used as state variables. Under the condition of satisfying the energy storage state of charge constraint and power balance constraint, the reference trajectory is used as the optimization objective to perform rolling optimization solution to obtain the optimal control sequence of water pump frequency and fertilizer channel opening.
[0072] Step S4: During the execution of the control sequence to drive the variable frequency water pump and the adjustable fertilizer channel, the actual photovoltaic power is detected in real time by the photovoltaic irradiance sensor. When the fluctuation of the actual photovoltaic power exceeds the preset threshold, the feedforward compensation amount is determined according to the preset mapping relationship with the photovoltaic power mutation amount as input and the water pump frequency compensation amount and the fertilizer channel opening compensation amount as output. The feedforward compensation amount is superimposed on the control sequence, and the variable frequency water pump and the adjustable fertilizer channel are continued to be driven by the superimposed control sequence.
[0073] Step S5: The actual concentration of the mixed fertilizer solution is detected in real time by the conductivity / pH sensor. Based on the deviation between the actual concentration and the target fertilizer solution concentration in the reference trajectory, the opening of the fertilization channel is adjusted accordingly.
[0074] This invention utilizes irradiance prediction data to obtain predicted photovoltaic power values for future periods, which are then used as input to a multivariate dynamic model. A model predictive control algorithm is employed to estimate future states in the prediction time domain and solve for the optimal control sequence in the control time domain. This proactive control strategy allows the system to adjust control parameters in advance of irradiance fluctuations, avoiding the lag inherent in pure feedback control.
[0075] This invention employs a feedforward-feedback composite control: The feedforward channel directly calculates and superimposes compensation amounts onto the control sequence based on a preset mapping relationship when a sudden change in actual photovoltaic power is detected. This adjustment is completed before concentration deviation occurs, effectively suppressing the impact of power disturbances on water and fertilizer output. The feedback channel continuously monitors the concentration of the mixed fertilizer solution and uses an adaptive PID controller to continuously correct the opening of the fertilization channel, eliminating residual deviations caused by model errors and unmeasurable disturbances. The feedforward and feedback channels operate in parallel, forming a composite control architecture of disturbance feedforward suppression + steady-state feedback correction.
[0076] This invention employs multivariate coupled optimization: a coupled model is established comprising a power balance sub-model, a variable frequency pump flow-frequency-power sub-model, and a water-fertilizer mixing dynamic sub-model. These three sub-models form a causal transmission chain of energy → flow rate → concentration through shared variables. During rolling optimization, the reference trajectory is used as the optimization objective, simultaneously satisfying SOC constraints, power balance constraints, flow rate constraints, and concentration constraints, thereby achieving coordinated optimization of energy scheduling and water-fertilizer ratio.
[0077] This invention employs adaptive mixing: adjustable guide vanes are installed in front of the static mixer, and the opening of the vanes is dynamically adjusted according to the real-time total flow rate and the target fertilizer concentration, so that the Reynolds number in the mixer is always maintained in the optimal range, ensuring the mixing uniformity under different flow conditions.
[0078] This method achieves its objective through the coordinated action of five steps: Step S1 acquires the current state of charge and predicted photovoltaic power, providing current state and future disturbance information for subsequent optimization; Step S2 generates reference trajectories for the target fertilizer concentration and target irrigation flow rate, providing a target benchmark for optimization; Step S3 performs rolling optimization based on a multivariate dynamic model, using the predicted photovoltaic power as model input and the reference trajectory as the optimization objective, comprehensively considering energy storage constraints and power balance constraints to obtain the optimal control sequence for pump frequency and fertilizer channel opening. This step is the core link in solving the multivariate coupling problem of energy-flow-fertilizer quantity; Step S4 detects sudden changes in photovoltaic power during execution, quickly determines the compensation amount through feedforward compensation mapping relationship and superimposes it onto the control sequence to achieve active suppression of disturbances; Step S5 performs feedback adjustment through concentration detection to further eliminate residual deviations, forming a composite control together with the feedforward compensation in S4. Through the closed-loop coordination of S1-S5, a complete control link from energy prediction → optimization decision-making → execution drive → feedforward disturbance rejection → feedback correction is realized, systematically solving the problems of stability and uniformity of water and fertilizer output under photovoltaic energy storage and supply conditions.
[0079] A photovoltaic array is a device that converts solar energy into direct current (DC) electricity. It consists of multiple photovoltaic modules connected in series and parallel and is installed in open areas of fields or on top of system equipment. The photovoltaic modules can be monocrystalline silicon photovoltaic panels, and the power of a single panel and the number of panels can be adjusted according to actual needs.
[0080] The energy storage unit stores excess electrical energy from the photovoltaic array and releases it when photovoltaic power is insufficient to ensure continuous power supply to the water and fertilizer head. The energy storage unit includes a battery bank and a bidirectional DC / DC converter. The battery bank uses lithium iron phosphate batteries. The bidirectional DC / DC converter is connected between the DC bus and the battery bank and controls the charging and discharging process of the batteries. When the photovoltaic output power is greater than the load power, the converter charges the excess energy into the battery; when the photovoltaic output power is less than the load power, the converter releases energy from the battery to supplement the DC bus. The charging and discharging power of the converter is regulated by the controller and constrained by the upper and lower limits of SOC. In this embodiment, the lower limit of SOC is... min Set to 20% (to prevent over-discharge from damaging the battery), SOC upper limit set to 95% (to reserve charging margin), normal operating range 30%-90%, and lower limit SOC threshold in low power protection mode. min It can be set to 25%, the recovery threshold SOC rec Set it to 40%.
[0081] The fertigation unit is the core component for integrated fertigation, responsible for mixing water and mother liquor in a target ratio before distributing the fertilizer solution to the field irrigation network. The fertigation unit includes a variable frequency pump, an adjustable fertilizer channel, a mother liquor container, and a static mixer. The variable frequency pump is the power source for the fertigation unit; the pump speed is adjusted via a frequency converter, thereby controlling the output flow rate. The mother liquor container stores high-concentration fertilizer mother liquor. The adjustable fertilizer channel connects the mother liquor container to the water supply pipeline and is used to regulate the mother liquor injection flow rate. The adjustable fertilizer channel includes a small fertilizer pump and a proportional control valve. The fertilizer pump draws mother liquor from the mother liquor container, and the proportional control valve receives opening commands from the controller, adjusting the mother liquor injection flow rate by changing the valve core opening.
[0082] The static mixer is located after the junction of the variable frequency water pump outlet and the adjustable fertilizer channel, and is used to mix the mother liquor with the water flow before outputting the fertilizer solution. The static mixer has multiple stages of fixed guide vanes inside, forcing the fluid to continuously change its flow direction within the pipe, thereby promoting the mixing of the mother liquor and the water flow. An adjustable guide vane is also installed at the inlet of the static mixer. The adjustable guide vane is electrically connected to a controller, which adjusts the flow based on the real-time total flow rate Q and the target fertilizer solution concentration C. refAdjusting the angle of the guide vanes alters the turbulence as the fluid enters the mixer. According to fluid mechanics principles, mixing efficiency is related to the Reynolds number (Re); excessively high or low Re values reduce mixing efficiency. By adjusting the vane opening, the Reynolds number within the mixer is maintained within a preset optimal range.
[0083] The sensor array is used to detect various physical quantities during system operation in real time, providing a data foundation for the controller's optimization and control decisions. The sensor array includes a photovoltaic irradiance sensor, an energy storage state-of-charge detection module, a conductivity / pH sensor, and an outlet flow meter, with optional inlet flow meter and mother liquor flow meter.
[0084] A photovoltaic irradiance sensor is mounted on the support frame housing the photovoltaic array, installed at the same angle as the photovoltaic modules, to detect solar irradiance in real time. The sensor's output signal is connected to the controller, which estimates the actual output power of the photovoltaic array based on the irradiance data. Furthermore, the controller compares the actual photovoltaic power with the predicted photovoltaic power value used in the rolling optimization solution to determine if a sudden change in photovoltaic power has occurred, serving as one of the conditions for triggering feedforward compensation. In this embodiment, a sudden change in photovoltaic power is determined to have occurred when the deviation between the actual photovoltaic power and the predicted value exceeds a preset threshold (e.g., 30% of the predicted value).
[0085] The energy storage state of charge (SOC) detection module is integrated into the battery management system of the battery pack. It is used to detect the remaining charge of the energy storage unit, i.e., the SOC. SOC is expressed as a percentage, with 100% indicating a fully charged battery and 0% indicating a fully discharged battery. The SOC signal is transmitted to the controller via a communication bus (such as RS485 or CAN bus) for initializing state variables in rolling optimization, determining constraints, and triggering low-power backup modes.
[0086] A conductivity / pH sensor is installed on the outlet pipe of the static mixer to detect the actual concentration of the mixed fertilizer solution. Since the fertilizer solution concentration is positively correlated with its conductivity (EC value), the sensor indirectly obtains the fertilizer solution concentration by measuring the conductivity. The sensor output signal is connected to the controller, which compares it with the target fertilizer solution concentration in the reference trajectory, calculates the concentration deviation, and uses it for feedback adjustment in step S5.
[0087] The outlet flow meter is installed on the outlet pipe of the static mixer to detect the real-time total flow rate Q (i.e., the total output flow rate of the water-fertilizer mixture) of the system. Since the mother liquor injection flow rate q is much smaller than the output flow rate Q of the variable frequency pump... water In practical applications, Q is approximately equal to Q waterThe outlet flow meter can be an electromagnetic flow meter or a turbine flow meter. In this embodiment, an electromagnetic flow meter is selected. The real-time total flow Q detected by the outlet flow meter has multiple uses: as one of the state variables for rolling optimization solution; as one of the input parameters for feedforward compensation mapping relationship; as the basis for adjusting the parameters of the adaptive PID controller; and as the basis for adjusting the opening of the adjustable guide vanes.
[0088] The inlet flow meter is installed at the outlet of the variable frequency water pump (i.e., upstream of the static mixer) to detect the output flow rate Q of the variable frequency water pump. water The mother liquor flow meter is installed at the outlet of the adjustable fertilizer channel to detect the mother liquor injection flow rate q. The signals from both flow meters are connected to the controller as input to a multivariate dynamic model, used to update the state variable values in the model in real time. The influent flow meter and mother liquor flow meter can be small-diameter electromagnetic flow meters, with their ranges matched to the flow ranges of the water pump output and mother liquor injection, respectively.
[0089] The controller is the core decision-making and control component of the system, responsible for acquiring sensor data, running control algorithms, and outputting control commands. The controller receives externally input irradiance prediction data (such as data obtained from a cloud platform via a 4G communication module) and internally executes the steps of the method of this invention.
[0090] In this embodiment of the invention, step S3 performs rolling optimization based on a multivariate dynamic model to obtain the optimal control sequence for the pump frequency and fertilizer channel opening. In this embodiment, step S3 is executed once per control cycle (e.g., 15 minutes). A multivariate dynamic model is pre-built within the controller, and this model is composed of the following coupled sub-models:
[0091] The power balance sub-model is used to describe the dynamic balance relationship between photovoltaic output power, energy storage charging and discharging power, and water pump power consumption.
[0092] P pv +P bat =P pump +P fert +P loss Among them, P pv P is the output power of the photovoltaic array. bat P represents the charging and discharging power of the energy storage unit, with discharging being positive and charging being negative. pump For the power consumption of the variable frequency water pump, P loss For the system power loss, P fert For adjustable fertilization channel power consumption, P bat Constrained by the upper and lower limits of SOC, in this embodiment the lower limit of SOC is 20% and the upper limit is 95%.
[0093] The variable frequency pump flow-frequency-power sub-model describes the output flow rate of the variable frequency pump after its frequency is converted through nonlinear efficiency. The mapping relationship between this output flow rate and the pump head is then used to determine the power consumption of the variable frequency pump.
[0094] Q water =η1(f)×f,P pump =η2(f)×Q water ×H; where Q water η1(f) is the output flow rate of the variable frequency pump, f is the frequency of the variable frequency pump, η2(f) is the nonlinear efficiency function of flow rate-frequency, η2(f) is the nonlinear efficiency function of power-flow rate, and H is the head of the variable frequency pump. The controller memory stores the flow rate-frequency curve and power-flow rate curve data tables obtained through bench test calibration.
[0095] The dynamic sub-model for water-fertilizer mixing describes the dynamic process of fertilizer concentration formed after mixing of the mother liquor injection flow rate and the variable frequency pump output flow rate. It consists of a static mixing equation connected in series with a first-order inertial and pure time-delay element. Its inputs are the mother liquor injection flow rate q(t) at time t and the variable frequency pump output flow rate Q at time t. water (t), the output is the actual fertilizer solution concentration C at the outlet of the static mixer at time t. out (t), the expression for the dynamic sub-model of water and fertilizer mixing is:
[0096] C(t) = q(t) × C m / (Q water (t)+q(t));
[0097] ;
[0098] Where C(t) is the theoretical concentration of the fertilizer solution after mixing at time t, C m Where is the concentration of the mother liquor, T is the mixing time constant, and τ is the lag time. For a moment The theoretical fertilizer solution concentration at that time;
[0099] P in the power balance sub-model pump and P fert The values of Q are determined by the control variables in the variable frequency pump flow-frequency-power sub-model and the water-fertilizer mixing dynamic sub-model, respectively; the output Q of the variable frequency pump flow-frequency-power sub-model is... water This serves as the input variable for the dynamic sub-model of water and fertilizer mixing.
[0100] In this embodiment of the invention, when the energy storage unit discharges, P bat When the energy storage unit is charging, P is positive. bat If negative, P lossConstrained by the upper and lower limits of the energy storage state of charge (SOC); in the step of performing rolling optimization, the optimization objective function includes a penalty term for deviation of the energy storage state of charge from the preset safe range, and the constraints include the upper and lower limits of the SOC, the maximum discharge power limit of the energy storage unit, the frequency range of the variable frequency water pump, the opening range of the adjustable fertilizer channel, the upper and lower limits of the system output flow detected by the outlet flow meter, the upper and lower limits of the outlet pressure, and the allowable fluctuation range of the concentration.
[0101] In this embodiment of the invention, the controller sets the optimization problem at the beginning of each control cycle as follows:
[0102] State variables: Current state of charge (SOC) (obtained from step S1), current concentration of the mixed fertilizer solution (C). out (Detected by conductivity / pH sensor), current total flow rate Q (detected by outlet flow meter). Optimization objective: Using the reference trajectory generated in step S2 as the optimization objective, such as the target fertilizer solution concentration C. ref =2.0 mS / cm, target irrigation flow rate Q ref =8.0 m³ / h. The optimization objective function is to minimize the weighted sum of squared deviations between the output value and the reference trajectory in the future prediction time domain, while adding a penalty term for deviations of the energy storage state of charge from a preset safe range (e.g., 30%-90%) to prevent the SOC from being too low or too high.
[0103] Constraints: SOC upper and lower limits (20%-95%), maximum discharge power limit (3kW), variable frequency pump frequency range (10-50Hz), adjustable fertilizer channel opening range (0-100%), system output flow upper and lower limits (2-10m³ / h), outlet pressure upper and lower limits (0.2-0.6MPa), and allowable concentration fluctuation range (±0.3mS / cm).
[0104] In this embodiment of the invention, the controller employs a model predictive control algorithm for rolling optimization, with the specific parameter settings as follows: control period T s =15 minutes, prediction time domain N p =8 (covering the next 2 hours), controlling the time domain N c =2 (covering the next 30 minutes). The prediction time domain covers at least one light fluctuation cycle (the typical light fluctuation cycle in this region is 0.5-2 hours), and the control time domain is smaller than the prediction time domain, so that the optimization can anticipate future light change trends while ensuring timely response to control commands.
[0105] In each control cycle, the controller performs the following calculation process:
[0106] The photovoltaic power prediction sequence (8 time points, one every 15 minutes) obtained in step S1 is used as the input of the multivariate dynamic model;
[0107] Use the current SOC, current concentration, and current flow rate as the initial values for the state variables;
[0108] Under the premise of satisfying all constraints, find the control sequence that minimizes the objective function;
[0109] The pump frequency commands f(k), f(k+1) and fertilizer channel opening commands u(k), u(k+1) for the next two control cycles are obtained.
[0110] Only the instructions f(k) and u(k) of the first control cycle are output to step S4 for execution, and the above process is repeated in the next control cycle.
[0111] In this embodiment of the invention, during the rolling optimization solution step, when it is predicted that the predicted photovoltaic power value will continue to be lower than the total power consumption of the variable frequency water pump and the adjustable fertilizer channel in the future period, and the current state of charge drops to the lower limit threshold SOC, the following steps are taken: min When the target irrigation flow rate in the reference trajectory is switched from the current value to the preset minimum flow rate value, the target fertilizer concentration in the reference trajectory is locked unchanged to enter the low power consumption minimum mode; when the predicted photovoltaic power recovers and the current state of charge rises to the recovery threshold SOC rec When this happens, exit the low-power backup mode.
[0112] In this embodiment of the invention, during the rolling optimization process, the controller simultaneously performs a conditional judgment on the low-power protection mode. When it is predicted that the predicted photovoltaic power value in the future period will be continuously lower than the total power consumption of the variable frequency water pump and the adjustable fertilizer channel, and the current state of charge (SOC) has dropped to the lower limit threshold, the protection mode is activated. min (In this embodiment, it is set to 25%), when the controller switches the target irrigation flow rate in the reference trajectory from 8.0 m³ / h to the preset minimum flow rate value of 4.0 m³ / h, and locks the target fertilizer concentration at 2.0 mS / cm to enter the low-power minimum flow mode. When the predicted photovoltaic power recovers and the state of charge rises to the recovery threshold SOC... rec (In this embodiment, it is set to 40%), then exit the minimum protection mode and restore the original reference trajectory.
[0113] In this embodiment of the invention, during the execution of steps S4 and S5, the method further includes: setting adjustable guide vanes before the static mixer, and adjusting the flow rate based on the real-time total flow rate Q detected by the outlet flow meter and the target fertilizer concentration C in the reference trajectory. ref Adjust the opening of the guide vanes to maintain the Reynolds number in the static mixer within a preset range.
[0114] In this embodiment of the invention, during the execution of the control sequence to drive the water and fertilizer head, step S4 detects photovoltaic power mutations in real time and performs feedforward compensation to actively suppress the impact of disturbances on water and fertilizer output before concentration deviation occurs.
[0115] The controller outputs the instruction for the first control cycle in the control sequence obtained in step S3 to the actuator: the water pump frequency instruction is sent to the frequency converter through the analog output module, and the frequency converter drives the variable frequency water pump to operate; the fertilizer channel opening instruction is sent to the proportional regulating valve through the analog output module, and the regulating valve adjusts the mother liquor injection flow to the corresponding opening.
[0116] In this embodiment of the invention, the real-time detection method for photovoltaic power surges is as follows:
[0117] During execution, the controller collects irradiance data every second using a photovoltaic irradiance sensor and converts the irradiance into actual photovoltaic power based on the rated parameters of the photovoltaic array. The controller then compares the actual photovoltaic power with the predicted photovoltaic power used in step S3 for optimization and calculates the deviation rate.
[0118] ;
[0119] In this embodiment, the preset threshold is 30%. When the deviation rate exceeds 30%, it is determined to be a sudden change in photovoltaic power. For example, if the predicted photovoltaic power is 1800W at a certain moment, but the actual photovoltaic power drops sharply to 1100W due to cloud cover, the deviation rate is 39%, which exceeds the threshold and triggers feedforward compensation.
[0120] Determination and superposition of feedforward compensation:
[0121] After triggering feedforward compensation, the controller reads a mapping relationship data table pre-established through offline calibration experiments from memory. This mapping relationship is expressed as the photovoltaic power mutation ΔP. pv The real-time total flow rate Q detected by the outlet flow meter is used as input, with the pump frequency compensation amount Δf as the variable. ff Compensation amount Δu for fertilization channel opening ff For output. The controller determines the output based on the current ΔP. pv Q is used to find the corresponding compensation value in the mapping data table through two-dimensional linear interpolation. The controller directly superimposes the compensation value onto the current control sequence, and the superimposed control command is immediately sent to the frequency converter and proportional control valve for execution. Due to the lag time of about a few seconds in the mixer, the above feedforward compensation is completed before the concentration deviation caused by the power change actually occurs, effectively suppressing the impact of disturbances on water and fertilizer output.
[0122] The offline calibration method can be as follows: In the offline calibration experiment, under multiple typical flow rates Q, step-type photovoltaic power mutations ΔP of different amplitudes are applied respectively.pv The optimal pump frequency compensation and fertilizer channel opening compensation are recorded to minimize fertilizer concentration fluctuations, and a two-dimensional lookup table relationship is formed by fitting the data using the least squares method.
[0123] In this embodiment of the invention, the mapping relationship is established through offline calibration experiments, and its input is the photovoltaic power mutation ΔP. pv The real-time total flow rate Q detected by the outlet flow meter is output as the pump frequency compensation amount Δf. ff Compensation amount Δu for fertilization channel opening ff The mapping relationship is stored in the memory of the controller.
[0124] In this embodiment of the invention, the rolling optimization solution employs a model predictive control algorithm, wherein the prediction time domain N p Covering at least one illumination fluctuation cycle, controlling the time domain N c Less than the prediction time domain N p .
[0125] In this embodiment of the invention, in step S2, the target fertilizer concentration and the target irrigation flow rate are dynamically adjusted based on the preset water and fertilizer plan for the current growth stage of the crop and the irradiance prediction data.
[0126] In this embodiment of the invention, in step S4, when a sudden change in photovoltaic power is detected, the water pump frequency compensation amount Δf... ff and the fertilizer application channel opening compensation amount Δu ff The control sequence is superimposed before the concentration deviation occurs; in step S5, the feedback adjustment is achieved through an adaptive PID controller, and the parameters of the adaptive PID controller are adjusted online according to the rate of change of the real-time total flow rate Q detected by the outlet flow meter and the deviation between the actual concentration and the target fertilizer concentration in the reference trajectory.
[0127] In this embodiment of the invention, the sensor group includes an influent flow meter and a mother liquor flow meter; in step S3, the multivariate dynamic model uses the influent flow rate detected by the influent flow meter and the mother liquor injection flow rate detected by the mother liquor flow meter as inputs to update the state variable values in the multivariate dynamic model.
[0128] In this embodiment of the invention, step S5 involves real-time detection of the actual concentration of the mixed fertilizer solution, and feedback adjustment based on the concentration deviation to eliminate the effects of residual deviation and model error after feedforward compensation.
[0129] A conductivity / pH sensor is installed on the outlet pipe of the static mixer. It collects the conductivity value of the fertilizer solution after one mixing at a preset sampling frequency and sends it to the controller via RS485 bus. The controller converts the conductivity value into the actual fertilizer solution concentration according to a pre-stored conductivity-concentration calibration curve.
[0130] The controller compares the actual concentration with the target fertilizer concentration in the reference trajectory and calculates the concentration deviation, which is calculated as: concentration deviation = target fertilizer concentration - actual concentration. At the same time, it calculates the rate of change of the deviation by dividing the difference between two consecutive sampled values by the sampling interval of 1 second, assuming that the current concentration is slightly higher than the target value.
[0131] The controller has a built-in adaptive PID controller with parameter K. p K i K d The adjustment is made online based on the real-time total flow rate Q and the concentration deviation change rate detected by the outlet flow meter. In this embodiment, the adjustment rules are as follows:
[0132] When the flow rate Q is large, the system response is fast, and a smaller K is used. p To avoid overshoot; when the flow rate is low, the system response is slow, so a larger K value is used. p To speed up the adjustment process.
[0133] When the rate of change of concentration deviation is large, increase K. d To suppress rapid changes; when the rate of change is small, decrease K. d To reduce noise sensitivity.
[0134] The adaptive PID controller is based on the concentration deviation e c and the current parameter K p K i K d Calculate the feedback correction amount Δu for the fertilization channel opening. fb :
[0135] ;
[0136] In this context, * represents multiplication, and ∫ represents integration. It is the integral variable.
[0137] in, ;
[0138] Where t is the current time; The target fertilizer concentration at time t in the reference trajectory; The actual fertilizer solution concentration at time t, as detected by the conductivity / pH sensor; Let t be the concentration deviation at time t.
[0139] The feedforward compensation in step S4 and the feedback adjustment in step S5 operate in parallel and work synergistically. The feedforward compensation completes the compensation superposition the instant the power change is detected (millisecond-level response), effectively suppressing most of the disturbances; the feedback adjustment continuously monitors the concentration deviation and corrects it in seconds-level cycles, eliminating the residuals and model errors of the feedforward compensation.
[0140] During the execution of steps S4 and S5, the controller synchronously adjusts the adjustable guide vanes. The controller adjusts the flow based on the real-time total flow rate Q detected by the outlet flow meter and the target fertilizer solution concentration C in the reference trajectory. ref The optimal opening of the guide vanes is calculated according to the preset control curve.
[0141] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0142] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply, characterized in that, This system includes a photovoltaic array, an energy storage unit, a water and fertilizer inlet, a sensor group, and a controller. The photovoltaic array and the energy storage unit jointly power the water and fertilizer inlet via DC bus coupling. The controller is electrically connected to both the sensor group and the water and fertilizer inlet. The water and fertilizer inlet includes at least a variable frequency water pump, an adjustable fertilizer channel, a mother liquor container, and a static mixer. The adjustable fertilizer channel is connected between the mother liquor container and the water supply pipe to adjust the mother liquor injection flow rate. The static mixer is located after the junction of the variable frequency water pump outlet and the adjustable fertilizer channel to mix the mother liquor with the water flow and output fertilizer solution. The sensor group includes at least: a photovoltaic irradiance sensor for detecting light intensity, an energy storage state-of-charge detection module for detecting the remaining power of the energy storage unit, a conductivity / pH sensor for detecting the concentration of the mixed fertilizer solution, and an outlet flow meter for detecting the system output flow rate; the controller receives externally input irradiance prediction data; the method includes the following steps: Step S1: Obtain the current state of charge of the energy storage unit through the energy storage state of charge detection module, and receive the irradiance prediction data for future time periods through the controller to determine the photovoltaic power prediction value; Step S2: Generate a reference trajectory for the target fertilizer solution concentration and target irrigation flow rate based on the crop's water and fertilizer requirements; Step S3: Based on the pre-established multivariate dynamic model, the photovoltaic power prediction value is used as the input of the multivariate dynamic model. The current state of charge, the current concentration of the mixed fertilizer solution detected by the conductivity / alkalinity sensor, and the current total flow rate detected by the outlet flow meter are used as state variables. Under the condition of satisfying the energy storage state of charge constraint and power balance constraint, the reference trajectory is used as the optimization objective to perform rolling optimization solution to obtain the optimal control sequence of water pump frequency and fertilizer channel opening. Step S4: During the execution of the control sequence to drive the variable frequency water pump and the adjustable fertilizer channel, the actual photovoltaic power is detected in real time by the photovoltaic irradiance sensor. When the fluctuation of the actual photovoltaic power exceeds the preset threshold, the feedforward compensation amount is determined according to the preset mapping relationship with the photovoltaic power mutation amount as input and the water pump frequency compensation amount and the fertilizer channel opening compensation amount as output. The feedforward compensation amount is superimposed on the control sequence, and the variable frequency water pump and the adjustable fertilizer channel are continued to be driven by the superimposed control sequence. Step S5: The actual concentration of the mixed fertilizer solution is detected in real time by the conductivity / pH sensor. Based on the deviation between the actual concentration and the target fertilizer solution concentration in the reference trajectory, the opening of the fertilization channel is adjusted accordingly.
2. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, The multivariate dynamic model is composed of the following coupled sub-models: The power balance sub-model is used to describe the dynamic balance relationship between photovoltaic output power, energy storage charging and discharging power, and water pump power consumption. P pv +P bat =P pump +P fert +P loss ; Among them, P pv P is the output power of the photovoltaic array. bat For the charging and discharging power of the energy storage unit, P pump For the power consumption of the variable frequency water pump, P loss For the system power loss, P fert Power consumption for adjustable fertilization channels; The variable frequency pump flow-frequency-power sub-model describes the output flow rate of the variable frequency pump after its frequency is converted through nonlinear efficiency. The mapping relationship between this output flow rate and the pump head is then used to determine the power consumption of the variable frequency pump. Q water =η1(f)×f,P pump =η2(f)×Q water ×H; where Q water η1(f) is the output flow rate of the variable frequency pump, f is the frequency of the variable frequency pump, η2(f) is the nonlinear efficiency function of flow rate-frequency, and H is the head of the variable frequency pump. The dynamic sub-model for water and fertilizer mixing takes the mother liquor injection flow rate q(t) at time t and the variable frequency pump output flow rate Q at time t as inputs. water (t), the output is the actual fertilizer solution concentration C at the outlet of the static mixer at time t. out (t), the expression for the dynamic sub-model of water and fertilizer mixing is: C(t)=q(t)×C m / (Q water (t)+q(t)); ; Where C(t) is the theoretical concentration of the fertilizer solution after mixing at time t, C m Where is the concentration of the mother liquor, T is the mixing time constant, and τ is the lag time. For a moment The theoretical fertilizer solution concentration at that time; P in the power balance sub-model pump and P fert The values of Q are determined by the control variables in the variable frequency pump flow-frequency-power sub-model and the water-fertilizer mixing dynamic sub-model, respectively; the output Q of the variable frequency pump flow-frequency-power sub-model is... water This serves as the input variable for the dynamic sub-model of water and fertilizer mixing.
3. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 2, characterized in that, When the energy storage unit discharges, P bat When the energy storage unit is charging, P is positive. bat If negative, P loss Constrained by the upper and lower limits of the energy storage state of charge (SOC); in the step of performing rolling optimization, the optimization objective function includes a penalty term for deviation of the energy storage state of charge from the preset safe range, and the constraints include the upper and lower limits of the SOC, the maximum discharge power limit of the energy storage unit, the frequency range of the variable frequency water pump, the opening range of the adjustable fertilizer channel, the upper and lower limits of the system output flow detected by the outlet flow meter, the upper and lower limits of the outlet pressure, and the allowable fluctuation range of the concentration.
4. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, In step S2, when it is predicted that the predicted photovoltaic power value will continue to be lower than the total power consumption of the variable frequency water pump and the adjustable fertilizer channel in the future period, and the current state of charge drops to the lower limit threshold SOC, min When the target irrigation flow rate in the reference trajectory is switched from the current value to a preset minimum flow rate value, and the target fertilizer concentration is locked unchanged, the reference trajectory in the low-power minimum mode is generated; when the predicted photovoltaic power recovers and the current state of charge rises to the recovery threshold SOC... rec When this happens, the low-power backup mode is exited, and the normal reference trajectory is generated again.
5. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, During the execution of steps S4 and S5, the method further includes: setting adjustable guide vanes before the static mixer, and adjusting the flow rate based on the real-time total flow rate Q detected by the outlet flow meter and the target fertilizer concentration C in the reference trajectory. ref Adjust the opening of the guide vanes to maintain the Reynolds number in the static mixer within a preset range.
6. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, The mapping relationship was established through offline calibration experiments, and its input is the photovoltaic power mutation ΔP. pv The real-time total flow rate Q detected by the outlet flow meter is output as the pump frequency compensation amount Δf. ff Compensation amount Δu for fertilization channel opening ff The mapping relationship is stored in the memory of the controller.
7. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, The rolling optimization solution employs a model predictive control algorithm, where the prediction time domain N... p Covering at least one illumination fluctuation cycle, controlling the time domain N c Less than the prediction time domain N p .
8. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, In step S2, the target fertilizer concentration and the target irrigation flow rate are dynamically adjusted based on the preset water and fertilizer plan for the current growth stage of the crop and the irradiance prediction data.
9. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 6, characterized in that, In step S4, when a sudden change in photovoltaic power is detected, the water pump frequency compensation amount Δf ff and the fertilizer application channel opening compensation amount Δu ff The control sequence is superimposed before the concentration deviation occurs; in step S5, the feedback adjustment is achieved through an adaptive PID controller, and the parameters of the adaptive PID controller are adjusted online according to the rate of change of the real-time total flow rate Q detected by the outlet flow meter and the deviation between the actual concentration and the target fertilizer concentration in the reference trajectory.
10. The water and fertilizer balanced supply control method based on photovoltaic-storage combined energy supply according to claim 1, characterized in that, The sensor group includes an influent flow meter and a mother liquor flow meter; in step S3, the multivariate dynamic model uses the influent flow rate detected by the influent flow meter and the mother liquor injection flow rate detected by the mother liquor flow meter as inputs to update the state variable values in the multivariate dynamic model.
