An intelligent water-saving irrigation system based on the Internet of Things

Abstract

The application belongs to the field of irrigation systems, and provides an intelligent water-saving irrigation system based on the Internet of Things, which comprises a control terminal, a data acquisition module and an irrigation execution module; the control terminal comprises a control module, a communication module, a data analysis module and a power module; the data acquisition module comprises a weather sensor, a soil humidity sensor and a plant growth state monitor electrically connected with the communication module; the irrigation execution module comprises a partition-controlled adjustable electric valve and a variable frequency water pump, and performs irrigation operation according to the instruction issued by the control module; the application can accurately perceive the soil humidity and the water requirement condition of plants by setting the control terminal, the data acquisition module and the irrigation execution module, the system can avoid unnecessary irrigation, effectively reduce water resource waste, realize full automation and intelligentization of the irrigation process, reduce manual intervention and improve work efficiency.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of irrigation systems, and particularly relates to an intelligent water-saving irrigation system based on the Internet of Things. BACKGROUND

[0002] Water resources are unevenly distributed worldwide, and many regions are facing serious water resource shortage problems. In the process of urban development, garden landscapes, as an important part of improving the quality of cities and ecological environment, need a large amount of water resources for irrigation and maintenance. With the influence of population growth, accelerated urbanization and climate change, the traditional extensive irrigation method has been difficult to meet the requirements of sustainable development, and more efficient and water-saving irrigation methods are urgently needed.

[0003] The rise of the Internet of Things technology enables the garden irrigation system to realize remote monitoring and management. Users can understand the running state of the irrigation system at any time and any place through mobile phones, computers and other terminal devices, and perform remote control.

[0004] Most of the existing garden irrigation methods rely on manual or traditional automatic control systems. The automation degree of the traditional irrigation system is low. The irrigation equipment is usually turned on and off regularly according to the preset time, which is difficult to adapt to the changeable weather conditions, and it is difficult to accurately perceive the real-time growth state of plants. The factors such as soil moisture and weather change are not considered for precise irrigation, resulting in waste of water resources or uneven irrigation.

[0005] Therefore, the technical personnel in the field propose an intelligent water-saving irrigation system based on the Internet of Things to solve the problems in the background technology. SUMMARY

[0006] In order to solve the above technical problems, the application provides an intelligent water-saving irrigation system based on the Internet of Things to solve the problems such as low automation degree of the traditional irrigation system in the prior art, easy to cause waste of water resources or uneven irrigation, etc.

[0007] An intelligent water-saving irrigation system based on the Internet of Things, comprising:

[0008] A control terminal comprising a control module, a communication module, a data analysis module and a power module;

[0009] A data acquisition module comprising a meteorological sensor, a soil moisture sensor and a plant growth state monitor electrically connected with the communication module; the data acquisition module continuously monitors meteorological conditions, soil moisture and plant growth conditions, and sends the collected data to the control module through the communication module; the control module calculates the irrigation demand based on the monitoring data of the data acquisition module;

[0010] An irrigation execution module, which comprises a zoning-controlled adjustable electric valve and a variable frequency water pump, and performs irrigation operation according to the instruction issued by the control module.

[0011] Preferably, the meteorological sensor in the data acquisition module monitors meteorological data in real time, including temperature, humidity, precipitation, wind speed and light intensity; and the soil moisture sensor detects the soil moisture content of the divided region.

[0012] Preferably, the control module is built-in with an irrigation strategy decision system, a reinforcement learning algorithm and a data processing unit of a control algorithm, and the control module automatically adjusts the irrigation time, frequency and water volume according to the water requirement characteristics of plants in different regions, current meteorological conditions and soil moisture data.

[0013] Preferably, the workflow of the reinforcement learning algorithm comprises:

[0014] Initializing a state space, which comprises meteorological parameters, soil moisture and plant growth state data;

[0015] Initializing an action space, which comprises a plurality of combined parameters of irrigation schemes, including irrigation time, irrigation frequency and water volume;

[0016] Setting a reward function, which is used to evaluate the effect of the system after executing the irrigation strategy, so as to optimize the irrigation scheme.

[0017] Preferably, the reward function is expressed as:

[0018] R=w1×f soil_humidity (S)+w2×f plant_growth (S)-w3×f water_usage (A);

[0019] Wherein, S represents the current state, A represents the action taken, f soil_humidity (S) is a function related to soil moisture, f plant_growth (S) is a function related to plant growth state, f water_usage (A) is a function related to irrigation water volume, and w1, w2 and w3 are weight coefficients.

[0020] Preferably, the control algorithm combines a genetic algorithm and a neural network for optimized control, wherein the genetic algorithm is used to optimize the structure and parameters of the neural network, and the connection weight, bias value and neuron number of the neural network are coded as chromosomes.

[0021] Preferably, a three-layer MLP neural network is expressed as:

[0022] y=f3(W3×f2(W2×f1(W1×x+b1))+b2)+b3;

[0023] where x is the input data, y is the output irrigation decision, W1, W2, W3 are the weight matrices of each layer, b1, b2, b3 are the bias vectors, and f1, f2, f3 are the activation functions.

[0024] Preferably, the intelligent irrigation scheduling algorithm comprises:

[0025] Collecting data: obtaining soil moisture, plant water demand, and irrigation priority information for different areas;

[0026] Calculating priority: calculating the irrigation priority of each area according to certain rules;

[0027] Sorting: sorting the areas according to the priority;

[0028] Developing an irrigation plan: allocating irrigation time and water to each area in order of priority from high to low;

[0029] Executing irrigation: controlling the irrigation execution module to perform irrigation operations according to the developed irrigation plan;

[0030] Monitoring and adjusting: monitoring the soil moisture changes and system operation status of each area in real time during the irrigation process, and adjusting the irrigation plan according to the actual situation.

[0031] Preferably, the power module includes a solar panel and a backup battery, the solar panel provides the main power for the system, and the backup battery automatically switches to power supply when the solar energy is insufficient.

[0032] Preferably, the communication module includes a wireless communication module and a cloud platform interface, the wireless communication module uses LoRa, NB-IoT low-power wide-area network technology; the cloud platform interface uploads the irrigation data to the cloud server in real time, and the user end remotely monitors and adjusts the irrigation parameters through the terminal device.

[0033] Compared with the prior art, the present application has the following beneficial effects:

[0034] The present application can accurately perceive soil moisture and plant water demand by setting the control terminal, data acquisition module and irrigation execution module, the system can avoid unnecessary irrigation, effectively reduce water waste, realize full automation and intelligentization of the irrigation process, reduce manual intervention, improve work efficiency; can provide personalized irrigation scheme according to the water demand characteristics of different plants, soil conditions and climate change, adapt to the needs of different gardens and crops. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 is a schematic diagram of the overall structure of the present application;

[0036] Figure 2A structural schematic diagram of a receipt collection module;

[0037] Figure 3 A structural schematic diagram of an irrigation execution module;

[0038] Figure 4 A layout schematic diagram of a green plant wetland. DETAILED DESCRIPTION

[0039] The embodiments of the present application will be further described in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present application, but cannot be used to limit the scope of the present application.

[0040] Example 1: As shown in the accompanying drawings, the present application provides an intelligent water-saving irrigation system based on the Internet of Things, comprising: a control terminal, the control terminal comprising a control module, a communication module, a data analysis module and a power module; Figure 1

[0041] A data collection module, the data collection module comprising a weather sensor, a soil moisture sensor and a plant growth state monitor electrically connected with the communication module; the data collection module continuously monitors weather conditions, soil moisture and plant growth conditions, and sends the collected data to the control module through the communication module; the control module calculates irrigation requirements based on the monitoring data of the data collection module;

[0042] An irrigation execution module, the irrigation execution module comprising a zoned control adjustable electric valve and a variable frequency water pump, and performing irrigation operation according to the instructions issued by the control module.

[0043] The weather sensor in the data collection module monitors real-time weather data, including temperature, humidity, precipitation, wind speed and light intensity; the soil moisture sensor detects the soil moisture content of the divided areas, forming a multi-point monitoring.

[0044] The weather sensor is used to monitor environmental temperature, humidity, wind speed, rainfall and light intensity, etc. This sensor can be connected with a weather station, so as to realize more accurate weather prediction;

[0045] The soil moisture sensor is used to monitor the water content in the soil of different areas, to ensure that the system makes corresponding irrigation adjustment according to the difference of soil moisture in different areas;

[0046] The plant growth state monitor is based on optical and non-contact measurement technology, which can monitor the chlorophyll content, leaf temperature and other growth state parameters of plants in real time, and comprehensively judge the water demand of plants.

[0047] The adjustable electric valve accurately adjusts the water flow according to the control instructions, supports multi-point irrigation control, and is suitable for the differentiated needs of different areas; the variable frequency water pump automatically adjusts the water pressure according to the irrigation requirements, reducing energy consumption. ​

[0048] The system can realize differentiated irrigation for different plant species or different areas through regional division, and ensure reasonable allocation of resources

[0049] The control module is built-in with an irrigation strategy decision system, a data processing unit of reinforcement learning algorithm and control algorithm. The control module automatically adjusts the irrigation time, frequency and water volume according to the water requirement characteristics of plants in different areas, current meteorological conditions and soil moisture data.

[0050] The workflow of the reinforcement learning algorithm includes:

[0051] Initialize the state space, which includes meteorological parameters, soil moisture and plant growth state data;

[0052] Initialize the action space, which includes multiple combination parameters of the irrigation scheme, including irrigation time, irrigation frequency and water volume;

[0053] Set the reward function, which is used to evaluate the effect of the system after executing the irrigation strategy, so as to optimize the irrigation scheme.

[0054] The reward function is represented as:

[0055] R=w1×f soil_humidity (S)+w2×f plant_growth (S)-w3×f water_usage (A);

[0056] Wherein, S represents the current state, A represents the action taken, f soil_humidity (S) is a function related to soil moisture, f plant_growth (S) is a function related to plant growth state, f water_usage (A) is a function related to irrigation water volume, and w1, w2 and w3 are weight coefficients.

[0057] The control algorithm combines genetic algorithm and neural network for optimization control, wherein the genetic algorithm is used to optimize the structure and parameters of the neural network, and the connection weight, bias value and neuron number of the neural network are coded as chromosomes.

[0058] The specific steps include: a randomly generated group of chromosomes represents different neural network structures and parameter combinations, forming an initial population;

[0059] Calculate the effect of each chromosome in processing irrigation data through fitness evaluation, and the evaluation indexes include irrigation effect and energy saving degree;

[0060] By selecting, crossing and mutating genetic operations, the structure and parameters of the neural network are optimized until the preset stopping condition is reached. After the neural network is optimized by the genetic algorithm, it can output accurate irrigation decisions based on the input meteorological data, soil moisture data and plant growth state data, including irrigation time, irrigation frequency and water quantity parameters. The neural network can use a multi-layer perceptron or a convolutional neural network structure.

[0061] The three-layer MLP neural network is represented as:

[0062] y = f3(W3 x f2(W2 x f1(W1 x x + b1)) + b2) + b3;

[0063] Where x is the input data, y is the output irrigation decision, W1, W2, W3 are the weight matrices of each layer, b1, b2, b3 are the bias vectors, and f1, f2, f3 are the activation functions.

[0064] The intelligent irrigation scheduling algorithm includes:

[0065] Collect data: Collect soil moisture, plant water demand, irrigation priority information in different areas;

[0066] Calculate priority: Calculate the irrigation priority of each area according to certain rules;

[0067] The difference between soil moisture and plant water demand is used as the main indicator of priority, and factors such as irrigation history and plant species are considered for adjustment;

[0068] Sort: Sort the regions according to the priority;

[0069] Develop an irrigation plan: According to the priority from high to low, allocate irrigation time and water quantity to each region in turn; In the allocation process, the total water resources limit and the system operation efficiency should be considered;

[0070] Execute irrigation: According to the irrigation plan, control the irrigation execution module to perform irrigation operation;

[0071] Monitoring and adjustment: During the irrigation process, real-time monitoring of soil moisture changes and system operation status in each region is carried out, and the irrigation plan is adjusted according to the actual situation.

[0072] Priority calculation formula:

[0073] Priority = w1 x (Plant_Water_Demand - Soil_Moisture) + w2 x f irrigation_history (R) + w3 x g plan ;

[0074] wherein, Plant_Water_Demand represents the plant water demand, Soil_Moisture represents the soil moisture, f irrigation_history (R) is an irrigation history related function, g planttype (T) is a plant species related function, w1, w2, w3 are weight coefficients.

[0075] The power module includes a solar panel and a backup battery, the solar panel provides the main power for the system, and the backup battery automatically switches power supply when the solar energy is insufficient.

[0076] The communication module includes a wireless communication module and a cloud platform interface, the wireless communication module adopts LoRa, NB-IoT low-power wide-area network technology; the cloud platform interface uploads the irrigation data to the cloud server in real time, the user end remotely monitors and adjusts the irrigation parameters through the terminal device, and the system also supports data docking with the third-party platform (such as the weather forecast platform) through API.

[0077] As can be seen from the above, when the system is working normally, the data acquisition module continuously monitors the environment and the plant growth conditions, and sends the collected data to the central control module through the wireless communication module;

[0078] The control module analyzes through the built-in AI algorithm, combines historical data and a preset plant growth model, and calculates the current optimal irrigation scheme in real time, and issues instructions to the irrigation execution module; the user can view the irrigation situation in real time through the mobile phone or computer terminal, and the system also supports manual adjustment of the irrigation scheme.

[0079] When the weather sensor detects that it will rain, the control module will automatically suspend irrigation to prevent over-irrigation; when the soil moisture is lower than the set value and the plant growth monitor judges that the plant needs water, the system will automatically start the irrigation equipment to ensure that the plant obtains sufficient water.

[0080] The system comprehensively considers multiple factors such as weather, soil moisture and plant growth state, can develop accurate irrigation strategies for different regions and plant species, avoids over-irrigation or insufficient irrigation, improves water resource utilization efficiency, and is beneficial to plant growth;

[0081] Through the cooperation of the reinforcement learning algorithm and the control algorithm combining genetic algorithm and neural network, the reinforcement learning algorithm comprehensively evaluates the irrigation effect through the reward function, and promotes the system to optimize the irrigation scheme; the genetic algorithm optimizes the neural network structure and parameters, so that the system can more accurately output reasonable irrigation decisions according to the input data, including time, frequency and water quantity, etc. parameters, which improves the scientificity and accuracy of irrigation decisions.

[0082] During the irrigation process, the data acquisition module continuously monitors the data, and the control module can adjust the irrigation strategy in a timely manner according to the real-time data (such as changes in weather, changes in plant growth state, etc.), better adapt to environmental changes, protect the plant growth environment, and at the same time avoid waste of water resources.

[0083] The variable frequency water pump automatically adjusts the water pressure according to the irrigation demand, reduces the energy consumption; the precise irrigation strategy also reduces the unnecessary waste of water resources, realizes the effect of energy saving and water saving, reduces the operation cost, and meets the requirements of sustainable development.

[0084] The communication module includes a wireless communication module and a cloud platform interface, adopts LoRa, NB-IoT low-power wide-area network technology, can upload the irrigation data to the cloud server in real time, and the user end adjusts the irrigation parameters through the terminal device. Remote monitoring, convenient management, improve the management efficiency and flexibility of the irrigation system.

[0085] The power module includes a solar panel and a backup battery, the solar panel provides main power for the system, the backup battery automatically switches power supply when the solar energy is insufficient, uses renewable energy, reduces the dependence on traditional energy, and improves the sustainability of the system

[0086] Example two:

[0087] In the plantation, the data acquisition module starts to work, the meteorological sensor monitors in real time that the summer temperature in this area often reaches above 35℃, the humidity is between 30%~40%, the precipitation is less, the wind speed is moderate, and the light intensity is strong. The soil moisture sensor detects that the soil moisture in different areas is quite different, the soil moisture in some areas is below 30% (the soil moisture suitable for fruit tree growth is set to 40%~60%), and the plant growth state monitor shows that the leaves of some fruit trees are slightly curled and the chlorophyll content has decreased, indicating that the fruit trees are facing certain water stress.

[0088] The reinforcement learning algorithm built-in the control module starts to work; first, initialize the state space, including meteorological parameters (temperature above 35℃, humidity 30%~40%, little precipitation, moderate wind speed, strong light intensity), soil moisture (below 30% in some areas), and plant growth state (leaf curling, chlorophyll content decreasing). The action space includes different irrigation scheme combinations, such as irrigation time (morning, evening, etc. Different time periods), irrigation frequency (once a day, twice a day, etc.), and water volume (determine different water volume levels according to fruit tree types and area size).

[0089] The reward function is set as R=ω1×f soil_humidity (S)+w2×f plant_growth (S)-w3×f water_usage(A); here set ω1=0.4, ω2=0.3, ω3=0.3.f soil_humidity (S) soil moisture related function, if the soil moisture can reach the appropriate range (40% ~ 60%) after irrigation, the function value is higher;f plant_growth (S) is a plant growth state related function, if the fruit tree leaf returns to normal, the chlorophyll content rises, the function value is higher;f water_usage (A) is an irrigation water related function, if the irrigation water is reasonable, it meets the needs of fruit trees and does not waste water resources, the function value is higher.

[0090] At the same time, the control algorithm combines genetic algorithm and neural network for optimization control. Genetic algorithm encodes the connection weight, bias value and neuron number of neural network as chromosome. For example, a set of chromosomes representing different neural network structures and parameter combinations are randomly generated to form an initial population. Through fitness evaluation, the effect of each chromosome in processing orchard irrigation data is calculated, and the evaluation index includes irrigation effect (whether the soil moisture and plant growth state can reach the ideal state) and energy saving degree (whether the water resources and energy are reasonably used).

[0091] Through genetic operations such as selection, crossover and mutation, the structure and parameters of neural network are optimized. After multiple iterations, the optimized neural network structure is obtained, which is represented by a three-layer MLP neural network as follows:

[0092] y = f3(W3 x f2(W2 x f1(W1 x x + b1)) + b2) + b3;

[0093] Where x is the input data (meteorological, soil moisture and plant growth state data), y is the output irrigation decision (irrigation time, frequency and water volume parameters).

[0094] According to the decision of the control module, the irrigation execution module starts to work; the adjustable electric valve of the partition control adjusts the opening degree according to the soil moisture and water demand of the fruit trees in different areas. For the area with soil moisture below 30%, the electric valve opening degree is larger to increase the water flow. The frequency conversion water pump automatically adjusts the water pressure according to the irrigation demand. Because the orchard area is large, a larger water pressure is needed to ensure that water can be delivered to each area.

[0095] The irrigation time is determined as 6 pm to 8 pm every day, the irrigation frequency is once a day, and the water volume is determined as 20 liters of water per fruit tree per irrigation (set value) according to the fruit tree species and area.

[0096] During the irrigation process, the data acquisition module continues to monitor weather, soil moisture, and plant growth status data, and the control module adjusts the irrigation strategy based on real-time data; if it rains suddenly during the irrigation process, the weather sensor detects that the precipitation reaches a certain standard, and the control module will stop irrigation in time and re-evaluate the subsequent irrigation demand.

[0097] Example Three:

[0098] The system is arranged in the garden landscape area of the urban park. The weather sensor detects that the air temperature in this area is between 15℃-25℃ in spring, the humidity is between 50%-60%, the precipitation is moderate, the wind speed is small, and the light intensity is moderate. The soil moisture sensor detects that the soil moisture in different areas also has differences, and the soil moisture in some areas is about 40%(the soil moisture suitable for flower growth is set to 45%-55%). The plant growth status monitor shows that the number of flower opening of some flowers is not as expected, and there may be a certain water supply problem.

[0099] The reinforcement learning algorithm initializes the state space, including weather parameters(temperature 15℃-25℃, humidity 50%-60%, moderate precipitation, small wind speed, moderate light intensity), soil moisture(partial area 40% or so), and plant growth status(flower opening number not as expected). The action space includes various combinations of irrigation schemes.

[0100] The reward function is set, ω1=0.3, ω2=0.4, ω3=0.3. soil_humidity (S) is the soil moisture related function, f plant_growth (S) is the plant growth status related function, f water_usage (A) is the irrigation water quantity related function.

[0101] The control algorithm combines genetic algorithm and neural network for optimization control. After similar genetic algorithm operation, the optimized neural network structure is obtained, which is used to output irrigation decision.

[0102] The irrigation execution module operates according to the decision of the control module; for the area with soil moisture of about 40%, the adjustable electric valve is properly opened to increase the water flow.

[0103] The frequency conversion water pump adjusts the water pressure to meet the irrigation demand. The irrigation time is determined as 9:00-11:00 am on Tuesday, Thursday and Saturday every week, the irrigation frequency is three times a week, and the water quantity is determined as 5 liters per square meter per irrigation(set value) according to the flower species and area.

[0104] During the irrigation process, the data is continuously monitored, and the control module adjusts the irrigation strategy according to the real-time data. For example, if it is found that the plant growth status in a certain area has improved significantly, the control module may appropriately reduce the subsequent irrigation water quantity or frequency of that area.

[0105] Example Four:

[0106] As shown in the accompanying Figure 4 The system is arranged in the greenery wetland, and the data acquisition module starts to work.

[0107] The meteorological sensor monitors that the temperature in the area is stable at 30°C in summer, the humidity is 50%, the precipitation accumulates to 20 mm in the past week, the average wind speed is 2 m / s, and the light intensity is 80000 lux.

[0108] The soil humidity sensor detects the specific conditions of the soil humidity in the four monitoring areas as follows:

[0109] The soil humidity of the part of the first monitoring area with Lycoris radiata as the base is 38%. The soil humidity range suitable for the growth of plants in this area is 42% to 52%. The soil humidity of the locations where the Narcissus tazetta, Sternbergia lutea, and Jonquil are arranged is slightly higher, being 39%.

[0110] The soil humidity of the part of the second monitoring area with Purple King Grass as the base is 39.5%. The soil humidity range suitable for the growth of plants in this area is 43% to 53%. The soil humidity of the areas where the Typha angustifolia, Cyperus alternifolius, and Small Sweetflag are arranged is relatively uniform.

[0111] The soil humidity of the part of the third monitoring area with Irish Pearl Grass as the base is 40%. The soil humidity suitable for the growth of flowers in this area is 44% to 54%. The soil humidity of the locations where the Hydrangea, Allium giganteum, Acorus gramineus, and Marsia are arranged is slightly higher than that of the Irish Pearl Grass, being 41%. The soil humidity of the parts where the Silver Leaf Chrysanthemum and the Molluscous Water-Snow are arranged is 40.5%.

[0112] The soil humidity of the part of the fourth monitoring area with Irish Pearl Grass as the base is 39%. The soil humidity suitable for the growth of plants in this area is 43% to 53%. The soil humidity of the location where the Solitary Rhododendron is arranged is 39.5%, and the soil humidity of the locations where the Shade Plant Hosta and the Aquatic Plant Vallisneria are arranged is 39.2%.

[0113] The plant growth state monitor displays:

[0114] The number of flowers of the Jonquil in the first monitoring area is 80% of the expected number, the degree of slight curling of the leaf edges of the parts of the Crinum asiaticum and the Nephrolepis auriculata reaches 5%, and the growth state of the Narcissus tazetta and the Sternbergia lutea is good.

[0115] The color of the leaves of the parts of the Typha angustifolia and the Cyperus alternifolius in the second monitoring area is 10% darker than the normal state.

[0116] The growth state of the Allium giganteum and the Acorus gramineus in the third monitoring area is 8% slower than the best state, and the leaves of the Silver Leaf Chrysanthemum and the Molluscous Water-Snow have a slight drooping sign.

[0117] The color intensity of the fourth monitoring area's azalea flowers is normal at 85%, and some leaves of the hosta, swordfish grass, and other plants have a slight yellowing area of 3% of the total leaf area.

[0118] The system combines past weather data to make predictions: The system connects to the weather database through the communication module to obtain the weather data of the region in previous summers; after data analysis, it is found that in the past five years, the average summer temperature in the region is 29-31°C, the average humidity is 48-52%, and the average precipitation is 80-100 mm per month. Based on these data and the current weather trend, the system predicts that in the next week, the temperature will remain between 29-31°C, the humidity may rise slightly to 52-55%, but the precipitation is expected to increase significantly.

[0119] The reinforcement learning algorithm built into the control module begins to work:

[0120] The state space is initialized, including meteorological parameters (temperature 30°C, humidity 50%, precipitation accumulates to 20 mm in the past week, average wind speed 2 m / s, light intensity 80,000 lux), soil moisture (soil moisture conditions in each area), and plant growth status (specific growth status of plants in each area).

[0121] The action space includes different irrigation scheme combinations, such as irrigation time (accurate to specific hours, such as 6 am, 7 am, etc.), irrigation frequency (specific to days, such as once every 1.5 days, etc.), and water volume (determined to the exact number of liters according to different plant species and area, such as 7.8 liters per square meter per irrigation in the first monitoring area, etc.).

[0122] The reward function is set as R(s, a) = ω1F(s) + ω2G(s) - ω3H(a); here ω1 = 0.32, ω2 = 0.33, ω3 = 0.35. The soil moisture related function F(s) is higher if the soil moisture in each area reaches the appropriate range after irrigation. For example, if the soil moisture in the first monitoring area reaches 42% or more, the function value starts to increase, and reaches a higher level when it reaches 50%. The plant growth status related function G(s) is higher if the leaves of plants in each area return to normal, the number of flowers increases, the color is bright, etc. For example, if the number of flowers of the long-lasting flower in the first monitoring area returns to more than 95% of the expected number, the leaf curling degree of bird's nest fern, acorn fern, and other plants decreases to less than 1%, the function value increases accordingly. The irrigation water volume related function H(a) is higher if the irrigation water volume is reasonable, meeting the plant's needs and not wasting water resources. For example, when the irrigation water volume is controlled within the range of the minimum water volume required for plant growth to 10% more than the required water volume, the function value is higher.

[0123] Meanwhile, the control algorithm combines genetic algorithm and neural network for optimization control. Genetic algorithm encodes the connection weights, bias values, and neuron numbers of neural network as chromosomes. For example, a set of chromosomes representing different neural network structures and parameter combinations are randomly generated to form an initial population. The effect of each chromosome in processing the wetland irrigation data of green plants is calculated through fitness evaluation, and the evaluation indexes include irrigation effect (whether the soil moisture and plant growth state can reach the ideal state) and energy saving degree (whether water resources and energy are reasonably utilized).

[0124] The irrigation execution module starts working according to the decision of the control module: for the areas with lower soil moisture, such as the first monitoring area, the second monitoring area, the third monitoring area, and the fourth monitoring area, the adjustable electric valve can adjust the opening degree according to different areas and plant species.

[0125] The first monitoring area can set the valve opening degree of the base part of the Pleione mixta to 45% according to the staggered structure of plants, and set the valve opening degree of the higher bird's eye, fern, etc. to 50%.

[0126] The second monitoring area sets the valve opening degree of the base part of the purple grass to 42% according to the transition requirements of the landscape, and sets the valve opening degree of the paper reed, umbrella grass, and small cattail to 45%.

[0127] The third monitoring area sets the valve opening degree of the base part of the Irish pearl grass to 43%, and sets the valve opening degree of the ball, large onion, calamus, and Mars flower to 48%, and sets the valve opening degree of the silver leaf chrysanthemum and the velvet marshmallow border to 45%.

[0128] The fourth monitoring area sets the valve opening degree of the Irish pearl grass around the center island to 42%, sets the valve opening degree of the azalea to 45%, and sets the valve opening degree of the hosta and the swordfish grass to 43%.

[0129] The frequency conversion water pump automatically adjusts the water pressure according to the irrigation requirements, taking into account the terrain and plant distribution of the green plant wetland to ensure that water can be uniformly delivered to each area. The water pressure is dynamically adjusted according to the requirements of different areas, such as 2.5 Pa for the first monitoring area, 2.3 Pa for the second monitoring area, 2.4 Pa for the third monitoring area, and 2.6 Pa for the fourth monitoring area.

[0130] The irrigation time is determined to be from 7:30 to 9:00 every morning, the irrigation frequency is once every 1.8 days, and the water quantity is determined according to the plant species and area of different areas. For example, the first monitoring area irrigates 7.8 liters of water per square meter each time; the second monitoring area irrigates 7.5 liters of water per square meter each time; the third monitoring area irrigates 7.6 liters of water per square meter each time; and the fourth monitoring area irrigates 8 liters of water per square meter each time.

[0131] During the irrigation process, the data acquisition module continues to monitor meteorological, soil moisture, and plant growth status data, and the control module adjusts the irrigation strategy based on real-time data. For example, if it suddenly rains during the irrigation process, and the meteorological sensor detects that the precipitation reaches 5 millimeters, the control module will stop irrigation in time and reevaluate the subsequent irrigation needs. If it is found that the plant growth status in a certain area has improved significantly, such as the number of open flowers of the first monitoring area reaching more than 90% of the expected value, the control module may appropriately reduce the subsequent irrigation water quantity or frequency of that area, adjusting the water quantity of each square meter to 7.5 liters per irrigation and the irrigation frequency to once every 2 days.

[0132] The embodiments of the present application are given for illustration and description only, although embodiments of the present application have been shown and described above, it is understood that the above-mentioned embodiments are exemplary and cannot be understood as limiting the present application, and those of ordinary skill in the art can make changes, modifications, replacements and deformations to the above-mentioned embodiments within the scope of the present application.

[0133] In the description of the present application, the terms "first" and "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features. The meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0134] In the present application, unless otherwise specifically defined and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the internal communication or interaction relationship of two elements. For those of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0135] In the present application, unless otherwise specifically defined and limited, the first feature is "on" or "under" the second feature, which can be direct contact between the first and second features, or indirect contact between the first and second features through an intermediate medium. Moreover, the first feature "above", "over" and "on" the second feature can be directly above or obliquely above the first feature, or only indicate that the horizontal height of the first feature is higher than that of the second feature. The first feature "below", "under" and "under" the second feature can be directly below or obliquely below the first feature, or only indicate that the horizontal height of the first feature is less than that of the second feature.

[0136] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are contained in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0137] In the drawings of the embodiments of the present application, only the structures related to the embodiments of the present disclosure are involved, and other structures can refer to the general design. In the case of no conflict, the same embodiment and different embodiments of the present application can be combined with each other.

[0138] Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments or make equivalent replacements for part of the technical features, and any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. An intelligent water-saving irrigation system based on the Internet of Things, characterized in that, include: The control terminal includes a control module, a communication module, a data analysis module, and a power supply module. The data acquisition module includes a meteorological sensor, a soil moisture sensor, and a plant growth status monitor, all electrically connected to the communication module. The data acquisition module continuously monitors meteorological conditions, soil moisture, and plant growth status, and transmits the collected data to the control module via the communication module. The control module calculates irrigation requirements based on the monitoring data from the data acquisition module. An irrigation execution module, comprising adjustable electric valves and variable frequency water pumps with zoned control, performs irrigation operations according to instructions issued by the control module; The control module has a built-in irrigation strategy decision-making system, reinforcement learning algorithm and control algorithm data processing unit. The control module automatically adjusts the irrigation time, frequency and water volume according to the water requirements of plants in different regions, current meteorological conditions and soil moisture data. The workflow of the reinforcement learning algorithm includes: Initialize the state space, which includes meteorological parameters, soil moisture, and plant growth status data; Initialize the action space, which includes multiple combined parameters of the irrigation scheme, including irrigation time, irrigation frequency and water volume; A reward function is defined to evaluate the effect of the system executing the irrigation strategy, in order to optimize the irrigation plan; the reward function is expressed as: ; in, Indicates the current state. Indicates the action taken. It is a function related to soil moisture. It is a function related to plant growth status. It is a function related to irrigation water volume. , , These are weighting coefficients; The control algorithm combines genetic algorithm and neural network for optimization control. The genetic algorithm is used to optimize the structure and parameters of the neural network, and the connection weights, bias values ​​and number of neurons of the neural network are encoded into chromosomes.

2. The intelligent water-saving irrigation system based on the Internet of Things as described in claim 1, characterized in that: The data acquisition module includes a meteorological sensor that monitors real-time meteorological data such as temperature, humidity, precipitation, wind speed, and light intensity; and a soil moisture sensor that detects the soil moisture content of the defined area.

3. The intelligent water-saving irrigation system based on the Internet of Things as described in claim 1, characterized in that: The three-layer MLP neural network is represented as follows: ; in, It is the input data. It is the output irrigation decision, It is the weight matrix of each layer. It is a bias vector. , It is an activation function.

4. The intelligent water-saving irrigation system based on the Internet of Things as described in claim 1, characterized in that: Intelligent irrigation scheduling algorithms include: Data collection: Obtain information on soil moisture, plant water requirements, and irrigation priorities in different regions; Priority calculation: Calculate the irrigation priority for each area according to certain rules; Sorting: Sort the regions according to their priority; Develop an irrigation plan: allocate irrigation time and water volume to each area in descending order of priority; Irrigation execution: Control the irrigation execution module to perform irrigation operations according to the established irrigation plan; Monitoring and Adjustment: During irrigation, monitor changes in soil moisture and system operation status in each area in real time, and adjust the irrigation plan according to the actual situation.

5. The intelligent water-saving irrigation system based on the Internet of Things as described in claim 1, characterized in that: The power module includes a solar panel and a backup battery. The solar panel provides the main power to the system, and the backup battery automatically switches to power supply when solar energy is insufficient.

6. The intelligent water-saving irrigation system based on the Internet of Things as described in claim 1, characterized in that: The communication module includes a wireless communication module and a cloud platform interface. The wireless communication module adopts LoRa and NB-IoT low-power wide area network technologies. The cloud platform interface uploads irrigation data to the cloud server in real time, and the user terminal can remotely monitor and adjust irrigation parameters through terminal devices.

Images