Fertigation control system based on the mariotte siphon | Scientific Reports

Feb 19, 2025

Scientific Reports volume 14, Article number: 23573 (2024) Cite this article

562 Accesses

Metrics details

This paper adopts the Mariotte siphon to simplify the nutrient solution preparation structure while improving the liquid mixing accuracy for nutrient solution. A liquid mixing model suitable for EC and pH regulation was constructed by combining fuzzy control algorithms, and a set of fertigation nutrient solution control equipment was designed and developed. During the experiment, comparison was made between the Fuzzy-PID algorithm and the traditional PID algorithm in terms of nutrient solution configuration. The results show that the Fuzzy-PID algorithm is smoother and more stable compared to the PID algorithm. Through analysis of the liquid mixing accuracy with the venturi type fertigation machine, it was found that the fertigation machine designed with the Mariotte structure is more accurate for liquid mixing, and can better meet the needs of crop growth.

With the rapid increase in global population, the demand for agricultural products and modern agricultural technology is increasing day by day1,2. In modern agriculture, crop cultivation using artificially prepared nutrient solutions is superior to traditional agriculture in terms of economics and environmental protection3,4. Nutrient solution can provide on-demand supply through irrigation management based on the actual growth needs of plants, while saving water resources and increasing yield and reducing environmental pollution, which is beneficial for sustainable agricultural production4,5,6,7,8.

At present, agricultural production mainly relies on integrated fertigation machine for automatic configuration of nutrient solutions, and the efficiency of nutrient solution use mainly depends on the management and design of integrated fertigation machine9.

Sun et al. developed and designed a multi-channel integrated fertigation machine using the Venturi suction method for the liquid mixing structure and a PLC as the core controller. Currently, high-precision nutrient solution can be configured in experimental environments10. Zhang et al. developed a fuzzy controller based on a microcontroller to quantitatively add solid fertilizers, stir and dissolve them before irrigation. Their developed fertigation control equipment is more convenient and concise compared to simple adoption of water-soluble fertilizers11. Pandey et al. developed a fertigation regulation equipment using a combination of metering pumps and PLC controllers. Through cultivation experiments, it was shown that using nutrient solution cultivation can improve yield and water use efficiency by 25–30% and 50%, respectively, compared to traditional cultivation methods12. Du et al. installed EC monitoring nodes in the field and adjusted fertigation irrigation management through changes in EC values. Through experimental comparison, it was found that using fertigation regulation management can reduce the average fertilizer usage by 10.89% compared to traditional fertilization13. ALEMAN et al. conducted fertigation on tomato seedling cultivation, and the experimental results showed that using nutrient solution control equipment can greatly enhance growth of tomato seedlings14. Wu et al. used a Venturi device as the fertilizer absorption structure and compared PID and fuzzy control algorithms. Through multiple sets of fertilizer absorption experiments, it was found that the fuzzy control algorithm has faster and more stable liquid mixing speed compared to traditional PID algorithms15. Neto et al. developed an automatic control system for tomato nutrient solution, which combines conductivity with environmental parameters to formulate corresponding irrigation management decisions. Through production experiments, this nutrient solution control system can minimize waste liquid emissions, help save fertigation resources, and reduce environmental pollution16. Zhao et al. successfully applied the integrated system of nutrient solution fertigation based on Mariotte siphons in watermelon cultivation17.

By studying the design and management methods of the current integrated fertigation machine, it can be concluded that the widely used fully automatic integrated fertigation machine can be divided into two types in terms of fertilizer absorption structure, the Venturi type and the fertilizer-injection pump type. The nutrient solution configuration algorithm is mainly based on PID and EC models to regulate and configure various elements in the nutrient solution18,19,20. When the fertilizer absorption structure adopts the Venturi type, its advantage is that it has a large amount of fertilizer absorption, but there are various factors influencing concentration of the nutrient solution, which leads to a large fertilizer absorption error. This type of fertilizer absorption structure is mainly suitable for large-scale production21,22,23,24. Compared to the Venturi type, the fertilizer injection pump permits smaller amount of fertilizer absorption and higher accuracy, making it suitable for small-scale production in greenhouses. However, due to its slow liquid mixing speed, the fertilizer injection pump often causes blockage and requires periodic maintenance.

This article focuses on the problems of large data fluctuations and low accuracy in the process of configuring nutrient solutions for current fertigation machine. In production applications, the accuracy of the solution preparation often fails to meet user requirements, and the unstable concentration of the nutrient solution not only fails to meet crop growth needs and has poor economic benefits, but also causes serious environmental pollution problems25,26. In response to the current problems in the application of fertigation machine, this article proposes using a mariotte siphon as a mother liquor storage tank, while utilizing the constant pressure and steady flow characteristics of the mariotte siphon to simplify the current liquid preparation structure and stably control its liquid preparation speed. In addition, to improve the accuracy of the liquid dispensing control algorithm, this article develops a set of fertigation machine that can be configured with high-precision nutrient solution based on the mariotte siphon using fuzzy control algorithm, so that it can better meet the production needs of on-site water and fertilizer nutrient solution and reduce application costs.

The structure and working principle of Venturi siphon are shown in Fig. 1. When using Venturi siphon, water flows from A1 end to A2 end. As water passes through d0 end, the cross-sectional area gradually narrows. According to Bernoulli’s equation, as the flow rate increases at d0, the P0 pressure will decrease, resulting in negative pressure. Due to the internal and external pressure difference between the Venturi throat and the liquid level in the stock tank, the Venturi siphon can be used to extract fertilizer from the stock tank.

Venturi siphon and installation schematic.

where Z1 is height at the venturi inlet (m), P1 is pressure at the output end (Pa), V1 is velocity at the Venturi inlet (m/s), Z0 is height at the venturi throat (m), P0 is pressure at the throat of the Venturi (Pa), V0 is velocity at the throat of the Venturi (m/s), γ is liquid specific weight (N/m3), g is acceleration due to gravity (m/s2), h is vertical height from the center of the venturi throat to the level of the fertilizer solution (m), h1-0 is water head loss due to changes in pipe diameter (m), ρ is solution density (kg/m³), Q0 is traffic on A0 end (m3/h).

From Eq. (2), it can be seen that when (h + P0/γ) ≥ 0 the Venturi suction is insufficient to effectively absorb fertilizer; When (h + P0/γ) < 0 it means that the Venturi siphon can effectively absorb fertilizers; From Eqs. (1) and (2), it can be seen that when using the Venturi siphon as the liquid absorption structure, the amount of fertilizer absorption depends on the liquid flow rate and pressure value at the input end of the Venturi siphon. When using a Venturi siphon for fertilizer mixing, fluctuations in the output flow rate and pressure of the water pump will cause corresponding fluctuations in the amount of fertilizer absorbed by the Venturi, thereby affecting accuracy of fertilizer mixing.

In response to the fluctuation and accuracy issues of the widely used Venturi siphon, this article adopts the Mariotte siphon to replace the Venturi siphon for improving dosing accuracy. The structure and cross-sectional view of the Mariotte siphon are shown in Fig. 2. The Mariotte siphon body is welded with stainless steel material and has a rectangular structure as a whole. At the top and bottom of the box, holes are opened simultaneously. A manual valve is installed above the box, mainly for adding high concentration nutrient solution, and an electromagnetic valve is installed below the box, mainly for adding corresponding nutrient solution to the mixing barrel. There is a vent hole in the lower right corner of the box, which can stabilize the pressure of the corresponding liquid level at atmospheric pressure according to the characteristics of the Mariotte siphon. When the internal liquid level of the Mariotte siphon is higher than the vent hole, and the liquid outlet below the box is opened, the flow rate of the outlet is only related to the height from the vent hole to the bottom discharge port of the solenoid valve. The corresponding outlet flow rate calculation equations are as follows:

where V is represents the flow rate of nutrient solution at the outlet of the Mariotte siphon box(m/s), g is represents the acceleration due to gravity (N/kg), H is represents height distance between the ventilation hole of the Mariotte siphon box and the outlet of the electromagnetic valve (m).

From the above equation, it can be seen that when the liquid level in the Mariotte siphon is higher than the vent hole, the flow rate at the outlet below the tank remains stable. When the flow rate at the outlet of the box is a fixed value, the amount of liquid to be added each time from the Mariotte siphon can be set by the on/off time of the lower solenoid valve.

Mariotte siphon structure and cross-sectional diagram of the box.

When configuring fertigation nutrient solution, first add concentrated nutrient solution from the stock tank to the mixing tank and stir it with clean water to achieve the target set value. After that, it is pumped through a water pump for field irrigation. Assuming the volume of the mixing tank is V, and since the concentration in the mixing tank varies dynamically and has a non-linear relationship with time, this article assumes that the current solution concentration is C(t). By using a variable frequency pump for liquid supply, it can be assumed that the liquid supply is constant at Q1, and the water supply in the mixing tank is Q2. When adding water, the ion concentration in the water is C1, and the concentration of the concentrated solution in the original tank is C2, Due to the fact that the amount of liquid added controlled by the electromagnetic valve during liquid dispensing is a time-dependent function, this article assumes that the amount of liquid added is Q(t). Based on the above analysis and the conservation of mass theorem, the following equation can be obtained:

where V is represents the volume of the mixing tank, C(t) is represents the current fertilizer concentration in the mixing tank, C1 is represents the concentration of fertilizer in the supplied water, C2 is represents the concentration of fertilizer in the stock tank, Q(t) is represents the amount of concentrated solution input into the mixing tank, Q1 is represents the supply rate of liquid, Q2 is represents the amount of water supplied to the mixing tank.

The left side of Eq. (4) shows the change in fertilizer mass in the mixing tank, while the right side shows the fertilizer mass obtained from the stock tank and the supplementary water, as well as the fertilizer output from irrigation. As the irrigation water supply has been pre-softened and filtered, the amount of fertilizer obtained from the stock tank and supplementary water can be adjusted accordingly. For the convenience of calculation, this article simplifies the model. Assuming that the ion concentration in the supply water after softening and filtration treatment is 0 g/mL, applying Laplace transformation to Eq. (2) can obtain the following equation,

In actual experiments, the volume of the mixing barrel V = 200 L, the supply volume Q1 = 320 L/h, and C2 = 240 g/L. Therefore, it can be inferred that the transfer function of the system mixing model is as follows.

PID (Proportional Integrated Derivative) control algorithm is a closed-loop control algorithm widely used in industrial control, automation systems, and robotics. It adjusts the output of the controller based on the error signal of the system, enabling the system to respond quickly and stably to the expected value. The PID control algorithm combines proportional, integral, and differential terms. The proportional term adjusts the controller output based on the current error size. It is proportional to the error and is used to quickly respond to system errors. Increasing the proportion coefficient (Kp) can increase the sensitivity of the system to errors, but an excessively large proportion coefficient may cause the system to overshoot or oscillate. The integral term adjusts the controller output based on the cumulative value of error over time. It is used to eliminate steady-state errors in the system and improve its static performance. The integration term can accumulate errors and continuously adjust the output, but a large integration coefficient (Ki) may cause the system to respond too slowly or produce integration saturation phenomenon. The derivative term adjusts the controller output based on the rate of error change. It is used to predict the future trend of errors and reduce the sensitivity of the system to changes in errors. Differential terms can improve system stability and suppress overshoot, but larger differential coefficients (Kd) may increase the sensitivity of the system to noise. Eq. (7) is the calculation equation for PID algorithm, and its logical block diagram is shown in Fig. 3:

where r(t) is setpoint value, y(t) is process variable or the actual value, e(t) is error value, which is the difference between the setpoint and the process variable, u(t) is PID controller output or control signal, Kp is proportional gain, Ki is integral gain, and Kd is Derivative gain.

Schematic diagram of gain scheduled PID controller.

Although the PID control algorithm is widely adopted in current industrial automatic control, it has some drawbacks. For example, the adjustment of Kp, Ki, Kd parameters relies heavily on engineering experience and trial and error processes. Due to varying demand for water, fertilizer, and EC/pH at different growth stages of crops, different combinations of Kp, Ki, Kd parameters are required for fertigation formulation under different operating conditions, and the mutual influence between parameters is complex. If the parameters are not adjusted properly, it may lead to unstable system response, overshoot or oscillation in the adjustment of fertigation EC/pH, and thus have a serious impact on on-site fertigation management. In response to such situations, it is necessary to incorporate advanced control algorithms such as fuzzy control and neural networks with traditional PID algorithms to overcome shortcomings of the traditional PID algorithm.

The fuzzy-PID control algorithm is a control strategy that combines fuzzy logic with PID control. It adjusts the Kp, Ki, and Kd parameters of the PID controller through the use of fuzzy logic. In traditional PID control, the selection of parameters usually relies on engineering experience, and the fuzzy PID control algorithm can adaptively adjust parameters based on the operating state of the system, enabling the controller to better adapt to uncertain and changing working conditions.

The fuzzy-PID control algorithm mainly includes 3 steps: (1) Fuzzi fication. Fuzzy the input actual error and error change rate, and map continuous actual values to the corresponding membership function. This can convert the input from the numerical domain to the form of a membership function, facilitating the subsequent fuzzy reasoning process. Define the fuzzy set of input variables e and ec as {NB, NM, NS, ZE, PS, PM, PB}. The corresponding domain is e, ec={-3,-2,-1,0,1,2,3}. Define the fuzzy set of output variables kp, ki, and kd as {NB, NM, NS, ZE, PS, PM, PB}. The corresponding domain ranges are kp=[-3 3], ki=[-3 3], kd=[-3 3]. Membership functions of input variables e and ec as show in Fig. 4.

Membership functions of input variables e and ec.

(2) Fuzzy inference. Use a set of fuzzy rules to derive the corresponding output membership function. Fuzzy rules are designed based on empirical knowledge and expert experience. The inference process utilizes the input fuzzy membership function and fuzzy rules to obtain a fuzzy output. The rule tables for kp, ki, and kd are shown in Tables 1, 2 and 3.

(3) Defuzzzi fication. Transform the fuzzy output obtained from fuzzy reasoning into actual control variables. The process of defuzzification maps the fuzzy output to the actual value, in order to adaptively adjust the Kp, Ki, and Kd parameters of the PID controller. The fuzzy resolution method adopts the centroid method. Membership functions of output variables kp, ki and kd as show in Fig. 5.

Membership functions of output variables kp, ki and kd.

Through the fuzzy-PID control algorithm, the parameters of the PID controller can be automatically adjusted according to the actual system state, thereby improving control performance and robustness. Eq. (8) is the calculation equation for fuzzy-PID algorithm, and its logical block diagram is shown in Fig. 6:

where r(t) represents the setpoint value, y(t) represents the process variable or target value, e(t) represents the error value, which is the difference between the setpoint and the target value, u(t) represents the PID controller’s control output, Kp is the PID proportional gain, Ki is the PID integral gain, Kd is the PID derivative gain, KpF is the fuzzy control PID proportional correction coefficient, KiF is the fuzzy control PID integral correction coefficient, KdF is the fuzzy control PID derivative correction coefficient.

Schematic diagram of fuzzy gain scheduled PID controller.

In order to verify the application effect of the fuzzy-PID control algorithm for nutrient solution preparation, this article uses MATLAB to conduct solution preparation simulation experiments on the afore-mentioned nutrient solution preparation model, PID, and fuzzy-PID algorithms. The EC regulation simulation results are shown in Fig. 7. During the simulation, the initial EC value was set to 1.5ms/cm. From the initial stage in Fig. 7, it can be observed that the fuzzy control PID has a smoother increase in data compared to the incremental PID algorithm. The maximum overshoot of the incremental PID algorithm and the fuzzy-PID algorithm are 0.12 and 0.04ms/cm, respectively. When the EC target value is set to 0ms/cm during simulation to simulate the sudden decrease of EC value in the mixing tank caused by the instantaneous addition of a large amount of clean water, the incremental PID algorithm and fuzzy-PID algorithm have a decrease of 1.03 and 0.87 ms/cm, respectively, to 0.47 and 0.63 ms/cm. Subsequently, when the EC target value recovers to 1.5 ms/cm, the control values of the incremental PID algorithm and fuzzy-PID algorithm begin to rise, Both the incremental PID algorithm and the fuzzy-PID algorithm exhibit overshoot again, with a maximum overshoot of 0.05 and 0.02ms/cm, respectively.

Comparison of incremental PID and Fuzzy-PID control for EC regulation.

The simulation results of pH regulation are shown in Fig. 8. During the simulation, the initial pH value was set to 6. From the initial control stage in Fig. 8, it can be observed that the fuzzy control PID has a greater decrease in data amplitude compared to the incremental PID algorithm. The minimum down regulation of the incremental PID algorithm and the fuzzy control PID algorithm are 5.93 and 5.87, respectively, and their maximum overshoot is 0.07 and 0.13, respectively. When the simulation data is stable, when the pH target value is set to 7 to simulate the sudden increase in pH value in the mixing tank caused by the instantaneous addition of a large amount of clean water, the incremental PID algorithm and fuzzy control PID algorithm have a maximum increase of 0.69 and 0.77, respectively, and a maximum upward appreciation of 6.69 and 6.77. Subsequently, when the pH target value returned to 6, the control values of the incremental PID algorithm and the fuzzy control PID algorithm gradually began to decrease. The control values of both the incremental PID algorithm and the fuzzy control PID algorithm showed overshoot again, with maximum overshoots of 0.03 and 0.09, and maximum downregulation values of 5.97 and 5.91, respectively.

Through the analysis of the comprehensive regulation effect of nutrient solution EC and pH in this article, it can be seen that the fuzzy-PID algorithm is more stable in terms of regulation compared to the traditional incremental PID algorithm.

Comparison of incremental PID and Fuzzy-PID control for pH regulation.

The processor of nutrient solution regulation equipment is mainly divided into upper computer and lower computer, and the upper computer uses Intel Core i7-4712MQ as a microcontroller. The processor has a main frequency of 2.3 GHz and a system memory of 8GB, which can meet the operational requirements of corresponding devices. Equipped with both Ethernet and USB interfaces, the processor can communicate directly with the lower computer by connecting to a 485 conversion module. The device can be remotely connected to the server directly through an Ethernet interface. The lower computer consists of a relay control board with 16 sets of DO outputs and two 485 communication modules, which can communicate data with the upper computer through the 485 interface. The lower computer controls the corresponding solenoid valves, liquid supply pumps, etc. through relay switches.

The nutrient solution control system is mainly composed of a data acquisition module, a liquid dispensing module, a liquid supply module, etc. in terms of functionality. The data acquisition module is mainly used to collect values such as EC and pH, liquid level, and supply flow in the mixing barrel, and feed back the detected data to the controller through the 485 bus. The volume design of the mixing tank is about 200 L. A liquid level sensor is installed inside the mixing tank. The control system analyzes the EC, pH, and liquid level in the current mixing tank. If the nutrient solution in the mixing tank does not meet the user’s required amount or concentration requirements for the next irrigation, the system automatically supplements the original solution and water volume by controlling the replenishment and replenishment solenoid valves according to the dispensing algorithm. When the concentration and amount of solution in the mixing tank meet the user’s set requirements, the system dispensing module stops working accordingly. When the equipment reaches the operating conditions for liquid supply, the equipment starts the liquid supply pump to transport the nutrient solution in the mixing tank to the corresponding irrigation area for crop growth irrigation. The system logic control structure is shown in Fig. 9.

Logic structure diagram of fertigation machine.

The nutrient solution configuration system is connected to the meteorological station through a 485 bus to obtain meteorological environment data within the greenhouse. Accumulated ETc in the greenhouse was calculated in real-time through the built-in algorithm of the system. When the accumulated ETc value of the crops reaches the set threshold, the system starts the irrigation pump to start irrigation. As the system begins to irrigate, the nutrient level in the mixing tank gradually decreases after irrigation. When the lowest liquid level is triggered by a decrease in liquid level, the system automatically starts liquid replenishment and simultaneously starts liquid dispensing. When the liquid level in the mixing tank rises to the highest set level at the top, the system automatically stops liquid replenishment. When the EC and pH sensors detect that the EC and pH values of the solution in the mixing tank are within the set target range, the system stops liquid dispensing. Otherwise, the system starts liquid dispensing. The integrated regulation equipment for nutrient solution fertigation is equipped with sensors such as EC, pH, liquid level, and flow meter to monitor the status of nutrient solution in real-time, enabling it to independently complete the three dynamic cycles of liquid preparation, supply, and replenishment in nutrient solution management, greatly reducing the workload of farmers.

Based on the logical structure design of the nutrient solution control equipment mentioned above, on-site debugging and installation of the fertigation machine have been completed in a greenhouse at the Beijing National Precision Agriculture Demonstration Station. On site installation uses four independent Mariotte siphons A, B, C, and D to hold the stock solution. The formula of stock solution components is shown in Table 4.

The tomato rock wool cultivation experiment was conducted using this type of fertigation machine in the solar greenhouse. The greenhouse was 40 m long and 6 m wide. The on-site test diagram is shown in Fig. 10. The fertigation nutrient solution control equipment will automatically calculate the crop ETc based on daily meteorological conditions and set the corresponding irrigation amount. Irrigation generally starts at 7 o’clock and ends at 18 o’clock according to daily meteorological conditions. Fig. 11 shows the liquid mixing effect diagram achieved using the Mariotte siphon based on PID and Fuzzy-PID algorithms during the production test process.

Fertigation machine and field experiment.

Nutrient solution control effect based on the mariotte structure.

From Fig. 11A, it can be seen that when the system EC is set to 2.5 ms/cm, the maximum overshoot value of EC regulation during the liquid dispensing process using the PID liquid dispensing algorithm is 2.93 ms/cm, and the maximum overshoot amount is 17.04%. The minimum offset value is 1.72 ms/cm, and the minimum overshoot is 31.04%. When using the Fuzzy PID liquid dispensing algorithm to regulate EC, the maximum overshoot value is 2.66 ms/cm, and the maximum overshoot amount is 6.36%. The minimum offset value is 1.85 ms/cm, and the minimum overshoot is 26.2%. From Fig. 11B, it can be seen that when the system pH is set to 6.0, the maximum overshoot value for pH adjustment during the dispensing process using the PID dispensing algorithm is 6.48, the maximum overshoot amount is 8%, the minimum offset value is 5.79, and the minimum overshoot is 3.5%. When using the Fuzzy-PID liquid dispensing algorithm to adjust pH, the maximum overshoot value is 6.4 and the maximum overshoot amount is 6.67%. The minimum offset value is 5.95, and the minimum overshoot is 0.83%. Through the above analysis, it can be concluded that the Fuzzy PID algorithm is more stable in regulating nutrient solution configuration compared to traditional PID algorithms.

During the experiment, a fertilizer mixing structure based on Venturi siphon and Mariotte siphon was used for liquid mixing experiments using the Fuzzy-PID algorithm. During the test, the EC was set to 2.5ms/cm and the pH was set to 6.0. The experimental results are shown in Fig. 12, the difference in EC value between Venturi and Mariotte configuration is relatively small, but the pH fluctuation range of the mixture when using a Venturi configuration is much larger than that when using a Mariotte structure configuration.

For the proposed fertigation mixture experiment using the Venturi and Mariotte mixture structure, the fertigation mixture sample data obtained through the experiment were statistically evaluated using mean absolute error (MAE), mean deviation (MBE), and root of mean square error (RMSE). The equation for calculating the sample statistical indicators is as follows:

where Pi and Qi is represent measured values and target values, respectively, and n represents the number of samples.

Comparison of EC and pH Effects Using Venturi and Mariotte Structures; (A) shows the effect of the Venturi siphon on EC; (B) shows the effect of the Mariotte structure on EC; (C) shows the effect of the Venturi siphon on pH; (D) shows the effect of the Mariotte structure on pH.

According to sample statistical analysis, it can be concluded that the closer the mean absolute error (MAE) and mean deviation (MBE) are to 0, the smaller the root of mean square error (RMSE), and the better the data calculation effect of the corresponding sample. Through calculation, it can be seen that when using the Venturi siphon configuration, MAE, MBE, and RMSE of EC are 0.06ms/cm, -0.01ms/cm, and 0.08ms/cm, respectively. When using the Mariotte structure configuration, MAE, MBE, and RMSE of EC are 0.04ms/cm, 0 ms/cm, and 0.06ms/cm, respectively. When using a Venturi siphon configuration, MAE, MBE, and RMSE of pH are 0.18, 0.11, and 0.23, respectively; When using the Mariotte structure configuration, the MAE, MBE, and RMSE of pH are 0.05, 0.04, and 0.07, respectively; The distribution diagram of its statistical indicators is shown in Fig. 13. When using the same liquid mixing algorithm, the fertigation machine using the Mariotte structure has higher liquid mixing accuracy and smaller liquid mixing fluctuation compared to the fertigation machine using the Venturi siphon.

From the statistical index distribution chart of the measured data of the water fertilizer mixture of Venturi and Mariotte in the following figure, we can see that the difference between the two liquid preparation structures is not particularly obvious when configuring EC, but the difference between the two liquid preparation structures is particularly significant when configuring pH. From this, it can be seen that when using the Venturi structure to configure pH values, its liquid stability is relatively poor compared to the Marriott structure. And the fluctuation of pH has a huge impact on crop growth, so it can be seen that the solution structure based on Marriott has a more obvious advantage in pH configuration compared to the Venturi structure solution structure.

Statistical distribution of measured data indicators for venturi and mariotte fertigation mixing.

This article uses a Mariotte siphon as the core dispensing structure for fertigation nutrient solutions. By utilizing its constant pressure and self-flow physical characteristics, high-precision dispensing requirements can be achieved through electromagnetic valve control alone. This simplifies the existing dispensing structure while greatly reducing equipment costs. At the same time, the Fuzzy-PID algorithm is used to adaptively adjust the coefficients of the control model, making it more suitable for liquid replenishment and dispensing compared to traditional PID algorithms, The dynamic circulation of the three supply fluids enables more stable operation. From February 23, 2023 to June 28, 2023, the equipment conducted a rockwool cultivation experiment of tomato in a solar greenhouse of the National Precision Agriculture Demonstration Station in Beijing. During the experiment, a total of 5 ears of tomato were harvested. A total of 1276.4 kg of tomatoes were harvested in the entire experimental area, with a total yield of 6.65 kg/m2 per unit area. Through production testing, it has been preliminarily proven that the integrated fertigation design scheme can achieve stable and reliable fully automatic operation.

However, through tomato production experiments, we found that although using Mariotte structured fertigation machine has better accuracy in fertigation management than Venturi style fertigation machine, due to Mariotte’s limited volume, multiple fertilizer supplements are often required during the experiment. However, due to the physical structure characteristics of Mariotte, it is more cumbersome compared to the Venturi type fertigation machine when replenishing the original solution. If Mariotte siphon-based fertigation machine is applied to small-scale crop cultivation, it is more suitable compared to the Venturi type fertigation machine.

Therefore, for small plant factories or other places that require high precision in fertigation, Mariotte siphon-based fertigation machine has more advantages compared to existing Venturi type fertigation machine. Due to the limited volume of Marriotte, multiple fertilizer supplements are often required during the experimental period. If further research is conducted on the structure of the Marriotte container solution supplement, it will better enhance and explore the application prospects of Mariotte siphon-based fertigation machine.

The data used to support the findings of this study are included within the article.

Khan, N. et al. Current progress and future prospects of agriculture technology: gateway to sustainable agriculture. Sustainability-Basel 13. https://doi.org/10.3390/su13094883 (2021).

Wei, Y. et al. Equipment and intelligent control system in aquaponics: a review. IEEE Access 7, 169306–169326. https://doi.org/10.1109/access.2019.2953491 (2019).

Article Google Scholar

Liu, R. et al. Chemical fertilizer pollution control using drip fertigation for conservation of water quality in Danjiangkou reservoir. Nutr. Cycl. Agrosyst. 98, 295–307. https://doi.org/10.1007/s10705-014-9612-2 (2014).

Article CAS Google Scholar

Valenzano, V., Parente, A., Serio, F. & Santamaria, P. Effect of growing system and cultivar on yield and water-use efficiency of greenhouse-grown tomato. J. Hortic. Sci. Biotechnol. 83, 71–75. https://doi.org/10.1080/14620316.2008.11512349 (2015).

Article Google Scholar

Wu, D. et al. Effect of different drip fertigation methods on maize yield, nutrient and water productivity in two-soils in Northeast China. Agric. Water Manag. 213, 200–211. https://doi.org/10.1016/j.agwat.2018.10.018 (2019).

Article Google Scholar

Nikolaou, G. et al. Energy and water related parameters in tomato and cucumber greenhouse crops in semiarid mediterranean regions. A review, part II: irrigation and fertigation. Horticulturae 7. https://doi.org/10.3390/horticulturae7120548 (2021).

Nikolaou, G., Neocleous, D., Christou, A., Kitta, E. & Katsoulas, N. Implementing sustainable irrigation in water-scarce regions under the impact of climate change. Agronomy 10. https://doi.org/10.3390/agronomy10081120 (2020).

Chojnacka, K. et al. A transition from conventional irrigation to fertigation with reclaimed wastewater: prospects and challenges. Renew. Sustain. Energy Rev. 130. https://doi.org/10.1016/j.rser.2020.109959 (2020).

Rodrguez, D., Reca, J., Martnez, J., Lpez-Luque, R. & Urrestarazu, M. Development of a new control algorithm for automatic irrigation scheduling in soilless culture. Appl. Math. Inf. Sci. 9, 47–56. https://doi.org/10.12785/amis/090107 (2015).

Article Google Scholar

Sun, G. et al. Design and testing of a nutrient mixing machine for greenhouse fertigation. Eng. Agric. Environ. Food 8, 114–121. https://doi.org/10.1016/j.eaef.2014.12.001 (2015).

Article Google Scholar

Zhang, Z., Chen, C., Li, H. & Xia, H. Design and evaluation of a control system for the fertigation device. J. ASABE 65, 1293–1302. https://doi.org/10.6041/j.issn.1000-1298.2022.06.040 (2022).

Article CAS Google Scholar

Pandey, K., Singh, K. G. & Singh, A. Multi-sensors based smart nutrient reuse management system for closed soilless culture under protected cultivation. Comput. Electron. Agr. 204 https://doi.org/10.1016/j.compag.2022.107495 (2023).

Du, C. et al. A cotton high-efficiency water-fertilizer control system using wireless sensor network for precision agriculture. Processes 9. https://doi.org/10.3390/pr9101693 (2021).

Aleman, C. C., Campos, F. B. & Mantovani, E. C. Precision tools for irrigation management of tomato seedlings. Biosci. J. 36. https://doi.org/10.14393/BJ-v36n2a2020-43252 (2020).

Wu, Y. et al. Optimal control algorithm of fertigation system in greenhouse based on EC model. Int. J. Agric. Biol. Eng. 12, 118–125. https://doi.org/10.25165/j.ijabe.20191203.4680 (2019).

Article Google Scholar

Steidle Neto, A. J. & Zolnier, S. De Carvalho Lopes, D. Development and evaluation of an automated system for fertigation control in soilless tomato production. Comput. Electron. Agr. 103, 17–25. https://doi.org/10.1016/j.compag.2014.02.001 (2014).

Article Google Scholar

Zhao, Y. et al. Study of nutrient solution management system based on mariotte device. Trans. Chin. Soc. Agric. Mach. 53, 380–387 (2022).

Google Scholar

Moon, T., Ahn, T. I. & Son, J. E. Forecasting root-zone electrical conductivity of nutrient solutions in closed-loop soilless cultures via a recurrent neural network using environmental and cultivation information. Front. Plant. Sci. 9. https://doi.org/10.3389/fpls.2018.00859 (2018).

Chiu, J. S. et al. Application of total internal reflection and heterodyne interferometry in electrical conductivity measurements. IEEE Sens. J. 16, 336–342. https://doi.org/10.1109/jsen.2015.2480887 (2016).

Article ADS CAS Google Scholar

Jeon, E. et al. Development of electrical conductivity measurement technology for key plant physiological information using microneedle sensor. J. Micromech. Microeng. 27. https://doi.org/10.1088/1361-6439/aa7362 (2017).

Ghassemi, H. & Fasih, H. F. Application of small size cavitating venturi as flow controller and flow meter. Flow Meas. Instrum. 22, 406–412. https://doi.org/10.1016/j.flowmeasinst.2011.05.001 (2011).

Article Google Scholar

Li, H. et al. Numerical and experimental study on the internal flow of the venturi injector. Processes 8. https://doi.org/10.3390/pr8010064 (2020).

Sharp, Z. B., Johnson, M. C. & Barfuss, S. L. Optimizing the ASME venturi recovery cone angle to minimize head loss. J. Hydraul. Eng. 144. https://doi.org/10.1061/(asce)hy.1943-7900.0001387 (2018).

Sobenko, L. R., Frizzone, J. A., Camargo, A. P., Saretta, E. & Rocha, H. S. d. characterization of venturi injector using dimensional analysis. Rev. Bras. Eng. Agr Amb. 23, 484–491. https://doi.org/10.1590/1807-1929/agriambi.v23n7p484-491 (2019).

Article Google Scholar

Liu, R., Zhang, Y., Ge, Y., Hu, W. & Sha, B. Precision regulation model of water and fertilizer for Alfalfa based on agriculture cyber-physical system. IEEE Access 8, 38501–38516. https://doi.org/10.1109/access.2020.2975672 (2020).

Article Google Scholar

Li, J., Meng, Y. & Li, B. Field evaluation of fertigation uniformity as affected by injector type and manufacturing variability of emitters. Irrig. Sci. 25, 117–125. https://doi.org/10.1007/s00271-006-0039-7 (2006).

Article CAS Google Scholar

Download references

This research was funded by the National Key Research and Development Program of China(2022YFD1900803), Major Achievement Cultivation Project of Beijing Academy of Agriculture and Forestry Sciences, Innovation Ability Construction Project of Beijing Academy of Agriculture and Forestry Sciences (KJCX20240407), China Agriculture Research System (CARS-03), Natural Science foundation of Henan Province (232300420105).

School of Agricultural Engineering, Jiangsu University, Zhenjiang, 212013, China

Wei Shi & Liping Chen

National Engineering Research Center of Intelligent Equipment for Agriculture, Beijing, 100097, China

Wei Shi, Xuzhang Xue, Wengang Zheng & Liping Chen

Department of Environmental Engineering, Yellow River Conservancy Technical Institute, Kaifeng, 475004, China

Feng Feng

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

X.X. and L.C. conceived the experiments, W.S. and F.F. conducted the experiment(s), W.Z. analysed the results. All authors reviewed the manuscript.

Correspondence to Xuzhang Xue or Liping Chen.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

Shi, W., Xue, X., Feng, F. et al. Fertigation control system based on the mariotte siphon. Sci Rep 14, 23573 (2024). https://doi.org/10.1038/s41598-024-75057-1

Download citation

Received: 25 January 2024

Accepted: 01 October 2024

Published: 09 October 2024

DOI: https://doi.org/10.1038/s41598-024-75057-1

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative