CN112488467A - Water planting crop fertilizer injection unit based on multiscale habitat information - Google Patents
Water planting crop fertilizer injection unit based on multiscale habitat information Download PDFInfo
- Publication number
- CN112488467A CN112488467A CN202011275810.XA CN202011275810A CN112488467A CN 112488467 A CN112488467 A CN 112488467A CN 202011275810 A CN202011275810 A CN 202011275810A CN 112488467 A CN112488467 A CN 112488467A
- Authority
- CN
- China
- Prior art keywords
- fertilizer
- information
- water
- fertilization
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000003337 fertilizer Substances 0.000 title claims abstract description 116
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 108
- 238000002347 injection Methods 0.000 title description 2
- 239000007924 injection Substances 0.000 title description 2
- 230000004720 fertilization Effects 0.000 claims abstract description 76
- 230000009471 action Effects 0.000 claims abstract description 43
- 230000007613 environmental effect Effects 0.000 claims abstract description 15
- 238000011161 development Methods 0.000 claims abstract description 13
- 230000002787 reinforcement Effects 0.000 claims abstract description 12
- 230000010365 information processing Effects 0.000 claims abstract description 8
- 239000007788 liquid Substances 0.000 claims description 18
- 238000000034 method Methods 0.000 claims description 17
- 235000015097 nutrients Nutrition 0.000 claims description 11
- 239000000243 solution Substances 0.000 claims description 10
- 230000006870 function Effects 0.000 claims description 8
- 239000003795 chemical substances by application Substances 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000005286 illumination Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 230000007704 transition Effects 0.000 claims description 6
- 238000010276 construction Methods 0.000 claims description 4
- 230000001276 controlling effect Effects 0.000 claims description 4
- 239000012452 mother liquor Substances 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 239000011259 mixed solution Substances 0.000 claims description 2
- 239000010413 mother solution Substances 0.000 claims description 2
- 230000002262 irrigation Effects 0.000 abstract description 7
- 238000003973 irrigation Methods 0.000 abstract description 7
- 230000008569 process Effects 0.000 description 8
- 230000018109 developmental process Effects 0.000 description 7
- 230000000875 corresponding effect Effects 0.000 description 6
- 238000001914 filtration Methods 0.000 description 5
- 238000003756 stirring Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 230000007774 longterm Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 210000003608 fece Anatomy 0.000 description 3
- 238000012840 feeding operation Methods 0.000 description 3
- 239000010871 livestock manure Substances 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000035784 germination Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 235000021049 nutrient content Nutrition 0.000 description 1
- 230000035790 physiological processes and functions Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01G—HORTICULTURE; CULTIVATION OF VEGETABLES, FLOWERS, RICE, FRUIT, VINES, HOPS OR SEAWEED; FORESTRY; WATERING
- A01G31/00—Soilless cultivation, e.g. hydroponics
- A01G31/02—Special apparatus therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Mining
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P60/00—Technologies relating to agriculture, livestock or agroalimentary industries
- Y02P60/20—Reduction of greenhouse gas [GHG] emissions in agriculture, e.g. CO2
- Y02P60/21—Dinitrogen oxide [N2O], e.g. using aquaponics, hydroponics or efficiency measures
Abstract
The invention discloses a water-cultured crop fertilizing device based on multi-scale habitat information, which comprises an information acquisition unit, an information processing unit, a fertilizing decision unit, a fertilizing control unit and a fertilizing machine, wherein the information acquisition unit is used for acquiring environmental factor information and crop growth and development information of water-cultured crops; the fertilization decision unit is used for acquiring water culture frame state information and fertilizer applicator state information, inputting the water culture frame state information and the fertilizer applicator state information into a fertilization intelligent decision control model established by using a depth reinforcement learning theory, and outputting an optimal regulation action; and the fertilization control unit is used for controlling the fertilizer applicator to execute complete regulation and control according to the optimal regulation and control action, wherein the crop growth period in the water culture frame state information is given by a water culture crop growth period discrimination model. The invention can accurately control the water and fertilizer feeding time, the water and fertilizer irrigation quantity and the proportion of the water and fertilizer in different growth periods, and solves the problem of dynamic control of the water and fertilizer according to the requirements of crops.
Description
Technical Field
The invention relates to a water-cultured crop fertilizing device based on multi-scale habitat information.
Background
In the process of water planting, scientific fertilization is carried out according to the nutrient content and the fertilizer requirement rule and characteristics of crop species, which is of great importance to the growth of crops.
Because the crop fertilization is a complex process, the fertilization effect is closely related to the crop growth environment and growth state, and the fertilization not only can affect the physiological process of crops, but also can affect the environment, such as: the irrigation time, irrigation quantity, temperature, humidity, illumination, CO2 concentration, leaf surface temperature and humidity, substance residue, leaf area increasing rate and other factors interact.
At present, the growth condition of the hydroponic crops is mainly monitored by environmental factors such as temperature, humidity, illumination, CO2 concentration and the like in the production of the hydroponic crops, and whether the short-term or long-term physiological and growth requirements of the crops are met is not known.
Meanwhile, the water and fertilizer integrated system is single in structure, water and nutrients are supplemented mostly through manual experience, the irrigation time and irrigation quantity of the water and fertilizer can not be controlled according to the crop growth environment and crop state information, the fertilization automation degree is low, the water and fertilizer can not be accurately regulated and controlled according to needs, the water and fertilizer cyclic utilization rate is low, and the loss is large.
How to reasonably, effectively and comprehensively introduce a plurality of information representing the growth environment and the growth state of crops into the water and fertilizer application, and the intelligent control of the water and fertilizer application is carried out on the basis of the information, thereby having important theoretical significance and application value for improving the production benefit of the water culture crops.
Disclosure of Invention
The invention aims to provide a water-cultured crop fertilizing device based on multi-scale habitat information, so as to realize rapid and accurate detection of water and fertilizer states of water-cultured crops and water and fertilizer putting decision, solve the problems of insufficient water and fertilizer putting and serious waste, and improve the utilization efficiency of water and fertilizer.
Therefore, the invention provides a water-cultured crop fertilizing device based on multi-scale habitat information, which comprises an information acquisition unit, an information processing unit, a fertilizing decision unit, a fertilizing control unit and a fertilizing machine, wherein the information acquisition unit is used for acquiring environmental factor information and crop growth and development information of water-cultured crops; the information processing unit is used for preprocessing the data acquired by the information acquisition unit; the fertilization decision unit is used for acquiring water culture frame state information needing water and fertilizer release and fertilizer applicator state information capable of normally executing water and fertilizer release operation, inputting the water culture frame state information and the fertilizer applicator state information into a fertilization intelligent decision control model established by utilizing a depth reinforcement learning theory, and outputting an optimal regulation action; and the fertilizing control unit is used for controlling the fertilizing machine to execute the regulation and control action after receiving the optimal regulation and control action, wherein the state information of the hydroponic frame comprises the growth period of the crops, and the growth period of the crops is given by a hydroponic crop growth period distinguishing model according to the growth and development information of the crops under different time scales.
The invention integrates a depth reinforcement learning theory and an irrigation fertilization amount control decision control technology, a hydroponic crop growth period discrimination model gives the growth period of each hydroponic frame cultivated crop according to environmental factor information and crop growth and development information, a real-time, short-term and long-term fertilizer requirement and environmental conditions of the growth period of the crop are given by an expert system, and a fertilization intelligent decision control model is provided by combining the hydroponic frame state information and the fertilizer applicator state information, so that the water and fertilizer feeding time, the water and fertilizer irrigation amount and the proportion of the hydroponic crops in different growth periods can be accurately controlled, the problem of dynamic control of water and fertilizer according to the crop requirements is solved, the limitation of fertilization control only according to the environmental factor information at present is overcome, and the fertilizer application according to need is realized, so the fertilizer consumption is greatly reduced, and the labor cost is reduced, the economic benefit is improved.
In addition to the objects, features and advantages described above, other objects, features and advantages of the present invention are also provided. The present invention will be described in further detail below with reference to the drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this application, illustrate embodiments of the invention and, together with the description, serve to explain the invention and not to limit the invention. In the drawings:
FIG. 1 is a block diagram of the construction of a hydroponic crop fertilization apparatus based on multi-scale habitat information in accordance with the present invention;
FIG. 2 is a flow chart of the construction of a fertilization intelligent decision control model of a hydroponic crop fertilization device based on multi-scale habitat information according to the present invention; and
fig. 3 is a workflow of a hydroponic crop fertilization apparatus based on multi-scale habitat information according to the present invention.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict. The present invention will be described in detail below with reference to the embodiments with reference to the attached drawings.
The invention provides a hydroponic crop fertilization device based on multi-scale habitat information, which is established by introducing growth environment and growth state information of hydroponic crops on the basis of a hydroponic crop growth monitoring device and comprises an information acquisition unit 10, an information processing unit 20, a fertilization decision unit 30, a fertilization control unit 40 and a fertilizer applicator 50.
The information acquisition unit comprises two information acquisition subunits: an environment factor information acquisition subunit 110 and a crop growth and development information acquisition subunit 120.
And after the two information acquisition subunits synthesize the data, the respective information is sent to the information acquisition unit in a unified standard format, and the acquisition of the growth process data of the hydroponic crops is completed.
The environment factor information acquisition subunit comprises an environment temperature sensor, an environment humidity sensor, an environment CO2 sensor, an environment illumination sensor, an EC sensor, a pH sensor, a liquid level sensor, a digital flowmeter and a pressure gauge. The environment temperature sensor, the environment humidity sensor, the environment illumination sensor and the environment CO2 sensor are installed in a greenhouse where the hydroponic crops are located, and are used for converting greenhouse temperature, humidity, illumination and CO2 concentration physical information into electrical information which is input into the information acquisition unit.
The liquid level sensor is arranged in the liquid mixing tank of each water culture frame and is used for judging the liquid level; EC sensor and pH sensor are used for detecting EC and pH value of nutrient solution, and digital flowmeter is used for detecting liquid manure input volume, and the manometer is used for judging pipeline pressure, and these sensors are equallyd divide and are installed respectively in the fertile pipeline of play fertilizer pipeline and each water planting frame that the mother liquor jar was located receives fertile pipeline.
The pH sensors are respectively arranged at positions away from the liquid level and the bottom of the mother liquid tank by a certain distance for obtaining the pH values of the two positions, and the uniformity degree of fertilizer stirring can be judged by comparing the difference values of the pH sensors, so that the rotating speed required by the stirring motor can be adjusted.
The crop growth and development information acquisition subunit comprises an image sensor for monitoring the height of the plant, the texture of the leaf and the color. For transparent hydroponic racks, the crop growth information acquisition subunit can also be used to monitor root morphology.
In order to carry out growth information acquisition to the crop in the vertical multilayer water planting frame including cylindricality water planting frame, A font water planting frame, image sensor can select for use the camera device that possesses all-round multi-level shooting ability, camera device supports automatic identification and pursuit monitoring object, supports intelligence to zoom and makes the monitoring object be in all the time and shoot the picture center to obtain the water planting crop high quality image of corresponding observation position and angle.
The information processing unit is used for processing the data acquired by the information acquisition unit by adopting different data preprocessing rules according to different data types in the crop growth process.
The data pre-processing rules include a uniform alignment of timestamps; denoising acquired data by adopting a finite amplitude filtering method, a median filtering method, an arithmetic mean filtering method or a wavelet threshold filtering method under the condition that a large amount of environmental noise and a plurality of abnormal data occur; formatting treatment is carried out on unformatted data including pictures, the pictures after conversion are verified, and an image processing algorithm is used, such as: coordinate transformation, image graying, image enhancement, image filtering and image segmentation, and then analyzing, processing and identifying the detection target characteristics.
The fertilization decision unit gives real-time, short-term and long-term fertilizer requirements and environmental conditions of different growth periods of crops through crop growth periods given by the hydroponic crop growth period discrimination model and an expert system established by integrating expert knowledge of greenhouse crop growth and cultivation and test results, and on the basis, a fertilization intelligent decision control model integrating environmental factors and crop growth and development information is established by utilizing a deep reinforcement learning theory.
The water culture crop growth period distinguishing model extracts image characteristics of crop images by acquiring growth and development data under different time scales, determines the corresponding relation between the image characteristics and the crop growth period, calculates the probability value of the crop growth period through a CNN algorithm according to the corresponding relation, and determines that the crop is in the growth period if the probability value of the growth period is greater than a preset value; otherwise, the uncertain growth cycle is labeled, and the algorithm of the labeled growth cycle is optimized when the steps of training and updating the algorithm model on line through the image database are executed. The types of the crop growth cycles stored in the algorithm model comprise: germination period, seedling throwing and growing period and fruiting period.
The construction of the intelligent fertilization decision control model specifically comprises the following steps:
and step S1, acquiring the state information of each water culture rack and the fertilizer applicator to form original information for deep reinforcement learning method training.
And step S2, establishing a fertilization intelligent decision control model by adopting a deep reinforcement learning theory, giving a reward and punishment value and state transition information to the reinforcement learning agent, determining the action space which can be selected by the agent and the value of the corresponding action, and determining the optimal regulation and control action according to the value. Wherein, the intelligent agent indicates the fertilization control unit, selects one or more from the water planting frame that awaits the fertilization and carries out liquid manure and put in the action.
And step S3, performing off-line training and learning by using the intelligent fertilization decision control model to obtain the trained intelligent fertilization decision control model.
And step S4, the water and fertilizer are put in by using the trained intelligent decision control model for fertilization.
In step S1, the original information includes hydroponic rack status information and fertilizer applicator status information:
the water culture rack state information is as follows: the number of the water culture frame, the type of crops, the growth period of the crops, the maximum capacity, the highest water level and the current water level of the liquid mixing tank, the maximum flow of an electromagnetic valve of a fertilizer receiving pipeline at the liquid mixing tank, the current EC and pH values of nutrient solution, the fertilizer requiring state and the putting state. Wherein the crop growth cycle is given by the hydroponic crop growth cycle discrimination model; the fertilizer requiring states are divided into normal, insufficient and excessive states, and are given by comparing and judging the current EC and pH value of the nutrient solution of the fertilizer outlet pipeline at the mother solution tank with the current EC and pH value of the nutrient solution of the fertilizer receiving pipeline at the mixed solution tanks of the water culture racks; the releasing state is divided into to-be-released state, releasing state and released state.
The state information of the fertilizer applicator is as follows: the number of the fertilizer applicator, the maximum flow of an electromagnetic valve of a fertilizer outlet pipeline at the mother liquor tank, the current EC and pH values of nutrient solution, the pressure value of the pipeline and the working state of the fertilizer applicator. The working state of the fertilizer applicator is divided into a distributed water and fertilizer feeding task, an unallocated water and fertilizer feeding task and an abnormity.
The intelligent fertilization decision control model comprises: an intelligent fertilization decision control environment model module and a value network module.
And the fertilization intelligent decision control environment model module gives a reward and punishment value and state transition information to the reinforcement learning intelligent body and determines an action space which can be selected by the intelligent body.
The value network module is used for abstracting the states of all the water culture frames and the fertilizer applicators in the greenhouse, outputting the values of the water culture frames and the fertilizer applicators corresponding to different actions in the states, selecting the optimal regulation and control action according to the value of the action corresponding to the abstracted states of the water culture frames and the fertilizer applicators, and feeding the selected optimal regulation and control action back to the intelligent fertilization decision control environment model; wherein the selected optimal action is in an action space provided by the fertilization intelligent decision control environment model.
The fertilization intelligent decision control environment model module comprises: a state transition unit, an action space unit, and a reward function unit.
The state transfer unit is used for implementing state transfer on the states of the water culture rack and the fertilizer applicator in the greenhouse at the current moment according to the state information of the water culture rack at the previous moment and the action of the current moment output by the value network module; the action space unit is used for determining an action range which can be selected by the intelligent agent according to the state information of the current greenhouse water culture frame and the state information of the fertilizer applicator; the reward function unit is used for calculating an output reward punishment value by utilizing a set reward function aiming at the water and fertilizer delivery regulation condition, and the output end of the reward function unit is connected with the input end of the value network; and the reward function is determined according to an optimization target of water and fertilizer delivery regulation.
The optimization target of water and fertilizer feeding regulation comprises the following steps: precision rate, punctuality rate. The precision rate refers to the degree of meeting the fertilizer using requirements given by an expert system, and the punctual rate refers to the degree of meeting the given water and fertilizer feeding time requirements.
The reward function is represented by the following equation: r is w 1F 1(a) + w 2F 2(B), wherein r is a reward and punishment value, F1(a) and F2(B) are respectively a water and fertilizer delivery precision rate and an on-time rate score, and w1 and w2 are weighted values.
The precision score is expressed as: f1(a) ═ log (1-a/N), a > 0; f1(a) ═ 1, and a ═ 0. And N is the total number of the water culture racks to be fertilized within the water and fertilizer feeding time period, and a is the number of the water culture racks to be fertilized with the water and fertilizer feeding precision error absolute value exceeding a given threshold value. The error of the water and fertilizer feeding precision refers to the difference between the water and fertilizer feeding quantity and the fertilizer consumption quantity given by an expert system.
The punctual rate score is expressed as: f2(B) ═ log (1-B/N), B > 0; f2(B) is 1 and B is 0. And N is the total number of the water culture racks to be fertilized within the water and fertilizer feeding time period, and b is the number of the water culture racks to be fertilized with the absolute value of the time error of water and fertilizer feeding exceeding a given threshold value. The error of the water and fertilizer feeding time requirement refers to the difference between the water and fertilizer feeding time and the given water and fertilizer feeding time.
The value network module is composed of a deep neural network, and selectable deep neural network models comprise network models such as ANN, CNN, RNN and LSTM and combinations or variants thereof, and the hydroponic frame state, the fertilizer applicator state and the value fitting are abstracted by using the deep neural network models.
In step S3, the method includes:
performing state abstraction according to the state information of the water culture frame and the state information of the fertilizer applicator to obtain the optimal regulation and control action in the current state, wherein the selectable action range of the value network is determined by an action space, and the optimal action selected by the value network is sent to the intelligent decision control model for fertilization; the intelligent fertilization decision control model is used for carrying out state transition according to the state of the water culture frame at the previous moment, the state of the fertilizer applicator and the action selection at the current moment, calculating a reward and punishment value according to a reward function, and feeding back the reward and punishment value and the changed state information to a value network; training and learning are continuously and iteratively carried out, and finally the fertilization intelligent decision control model is converged.
In step S4, the method includes:
the method comprises the following steps that a fertilization control unit firstly outputs state information of all water culture racks and a fertilizer applicator in a greenhouse at the current moment according to an actual real water and fertilizer release regulation and control environment, and transmits the state information to a fertilization intelligent decision control model, and the fertilization intelligent decision control model outputs an optimal regulation and control action according to input state information; returning the optimal regulation and control action to the fertilization control unit; the fertilization control unit receives the optimal regulation action, controls the fertilizer applicator to execute the regulation action, and then when the monitoring waiting time reaches the set water and fertilizer release interval, the fertilization control unit continues to release the regulation and control environment according to the current water and fertilizer, sends all the water culture racks and the state information of the fertilizer applicator in the greenhouse at the current moment to the intelligent fertilization decision control model, and acquires new optimal regulation and control action, so that the process is continuously circulated, and finally all the water culture racks are used for cultivating crops to be fertilized.
The process of outputting the optimal regulation and control action according to the input state information further comprises the following steps:
and a value network module in the intelligent fertilization decision control model obtains the optimal regulation and control action in the current state according to the state information input by the fertilization control unit and by combining the action range and the constraint condition provided by the action space. And the constraint conditions comprise the maximum capacity of the liquid mixing tank, the highest water level and the maximum flow of the electromagnetic valve.
The fertilization control unit is used for controlling the fertilizer applicator to execute water and fertilizer feeding operation according to the fertilization intelligent decision control model, and guiding or reminding a user to adjust the fertilization proportion and the fertilization flow.
The fertilization control unit can also be used for controlling environment regulation and control equipment including a fan and a wet curtain in the greenhouse to execute environment regulation and control operation in the greenhouse, so that the current environmental factor information reaches the environmental conditions given by the expert system.
The fertilization control unit can also judge the uniformity degree of fertilizer stirring according to the comparison of the difference values of the two pH sensors installed in the mother liquid tank of the fertilizer applicator, then control the rotating speed required by the stirring motor through the PLC, and automatically start and stop the stirring motor.
The fertilizer distributor mainly comprises inlet trunk line, filter, intake pump, inlet solenoid valve, play fertile pipeline, fertilization solenoid valve, agitator motor, mother liquor jar, fertilization pump, play fertile pipe, combines to obtain water planting frame state information with fertilizer distributor state information, by fertilization the control unit carries out PLC control to it, makes each water planting frame liquid manure input volume of waiting to fertilize reach the volume of using fertilizer that expert system gave.
With reference to fig. 3, the water planting crop fertilizing device of the invention has the following working flow:
firstly, acquiring growth process data of hydroponic crops through an information acquisition unit, processing the data by an information processing unit, inputting the processed data into a fertilization decision unit, giving a growth period of each hydroponic frame cultivated crop through a hydroponic crop growth period discrimination model, and giving real-time, short-term and long-term fertilizer requirements and environmental conditions of the growth period of the crop through an expert system.
And secondly, acquiring the state information of the water culture frame needing to be subjected to water and fertilizer feeding and the state information of the fertilizer applicator capable of normally executing water and fertilizer feeding operation.
And then inputting the state information of the water culture frame needing to carry out water and fertilizer feeding and the state information of a fertilizer applicator capable of normally executing water and fertilizer feeding operation into the intelligent fertilization decision control model, and giving out an optimal regulation action.
And finally, the fertilization control unit receives the optimal regulation action and controls the fertilizer applicator to execute the regulation action. When monitoring the waiting time and reaching the set water and fertilizer feeding interval later, the fertilization control unit continues to feed the regulation and control environment according to the current water and fertilizer, sends the state information of all the water culture racks and the fertilizer applicators in the greenhouse at the current moment to the intelligent fertilization decision control model, and acquires new optimal regulation and control actions, so that the processes are continuously circulated, and finally all the water culture racks are used for cultivating crops to be fertilized.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (5)
1. A water planting crop fertilizing device based on multi-scale habitat information is characterized by comprising an information acquisition unit, an information processing unit, a fertilizing decision unit, a fertilizing control unit and a fertilizing machine,
the information acquisition unit is used for acquiring environmental factor information and crop growth and development information of the hydroponic crops;
the information processing unit is used for preprocessing the data acquired by the information acquisition unit;
the fertilization decision unit is used for acquiring water culture frame state information needing water and fertilizer release and fertilizer applicator state information capable of normally executing water and fertilizer release operation, inputting the water culture frame state information and the fertilizer applicator state information into a fertilization intelligent decision control model established by utilizing a depth reinforcement learning theory, and outputting an optimal regulation action;
the fertilizing control unit is used for controlling the fertilizing machine to execute the regulating action after receiving the optimal regulating action,
the hydroponic frame state information comprises a crop growth period, wherein the crop growth period is given by a hydroponic crop growth period distinguishing model according to crop growth and development information under different time scales.
2. The multi-scale habitat information-based hydroponic crop fertilization device of claim 1, wherein the information acquisition unit comprises an environmental factor information acquisition subunit and a crop growth and development information acquisition subunit, the environmental factor information acquisition subunit is used for acquiring greenhouse temperature, humidity, illumination and CO2 concentration in a greenhouse in which hydroponic crops are located, and is also used for acquiring liquid level in a liquid mixing tank of a hydroponic rack, EC and pH values of nutrient solution in the liquid mixing tank, water and fertilizer feeding amount and pipeline pressure, the crop growth and development information acquisition subunit comprises an image sensor for monitoring plant height, leaf texture and color, wherein, for the transparent hydroponic rack, the crop growth and development information acquisition subunit can also be used for monitoring root morphology.
3. The multi-scale habitat information-based hydroponic crop fertilization device of claim 1, wherein the construction of the fertilization intelligent decision control model comprises the following steps:
step S1, acquiring the state information of each water culture frame and the state information of the fertilizer applicator to form original information for deep reinforcement learning method training;
step S2, establishing a fertilization intelligent decision control model by adopting a deep reinforcement learning theory, giving a reward and punishment value and state transition information to a reinforcement learning intelligent agent, determining an action space which can be selected by the intelligent agent and a value of an action corresponding to the action space, and determining an optimal regulation and control action according to the value, wherein the intelligent agent refers to a fertilization control unit;
and step S3, performing off-line training and learning by using the intelligent fertilization decision control model to obtain the trained intelligent fertilization decision control model.
4. The multi-scale habitat information-based hydroponic crop fertilization device of claim 3,
the water culture rack state information is as follows: the number of the water culture frame, the type of crops, the growth period of the crops, the maximum capacity, the highest water level and the current water level of the liquid mixing tank, the maximum flow of an electromagnetic valve of a fertilizer receiving pipeline at the liquid mixing tank, the current EC and pH values of nutrient solution, the fertilizer requiring state and the putting state; wherein the crop growth cycle is given by the hydroponic crop growth cycle discrimination model; the fertilizer requiring states are divided into normal, insufficient and excessive states, and are given by comparing and judging the current EC and pH value of the nutrient solution of the fertilizer outlet pipeline at the mother solution tank with the current EC and pH value of the nutrient solution of the fertilizer receiving pipeline at the mixed solution tanks of the water culture racks; the releasing state is divided into to-be-released state, releasing state and released state.
The state information of the fertilizer applicator is as follows: the number of the fertilizer applicator, the maximum flow of an electromagnetic valve of a fertilizer outlet pipeline at the mother liquor tank, the current EC and pH values of nutrient solution, the pressure value of the pipeline and the working state of the fertilizer applicator. The working state of the fertilizer applicator is divided into a distributed water and fertilizer feeding task, an unallocated water and fertilizer feeding task and an abnormity.
5. The multi-scale habitat information based hydroponic crop fertilization device of claim 3, wherein the fertilization intelligence decision control model comprises: the intelligent fertilization decision control environment model module comprises a fertilization intelligent decision control environment model module and a value network module, wherein the fertilization intelligent decision control environment model module comprises: the system comprises a state transition unit, an action space unit and a reward function unit, wherein the value network module is composed of a deep neural network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011275810.XA CN112488467A (en) | 2020-11-16 | 2020-11-16 | Water planting crop fertilizer injection unit based on multiscale habitat information |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011275810.XA CN112488467A (en) | 2020-11-16 | 2020-11-16 | Water planting crop fertilizer injection unit based on multiscale habitat information |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112488467A true CN112488467A (en) | 2021-03-12 |
Family
ID=74930394
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011275810.XA Pending CN112488467A (en) | 2020-11-16 | 2020-11-16 | Water planting crop fertilizer injection unit based on multiscale habitat information |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112488467A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114662609A (en) * | 2022-04-01 | 2022-06-24 | 电子科技大学成都学院 | Intelligent greenhouse farm management method and system |
CN117437080A (en) * | 2023-12-20 | 2024-01-23 | 潍坊汇金海物联网技术有限公司 | Farmland soil water and fertilizer nutrition supplementing system and method |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106707767A (en) * | 2017-03-13 | 2017-05-24 | 山东农业大学 | System and method for integrally and intelligently controlling water and fertilizer in field based on multi-source information fusion |
CN108762079A (en) * | 2018-06-04 | 2018-11-06 | 河海大学常州校区 | The traversing process control system and method for cutter suction dredger based on deeply study |
CN109511349A (en) * | 2018-12-20 | 2019-03-26 | 山东锋士信息技术有限公司 | A kind of intelligent accurate fertigation device and control method for realizing that nutrient matches in real time |
CN109673480A (en) * | 2019-02-22 | 2019-04-26 | 福州阿里他巴信息科技有限公司 | Intelligent water and fertilizer integrated sprinkling irrigation system and control method |
CN110187688A (en) * | 2019-06-14 | 2019-08-30 | 青岛农业大学 | Industrialized agriculture intelligence liquid manure integral control system and control method |
CN110214506A (en) * | 2019-06-26 | 2019-09-10 | 北京农业智能装备技术研究中心 | Liquid manure management-control method and system |
CN110246051A (en) * | 2019-05-23 | 2019-09-17 | 成都信息工程大学 | The method that crop cycle model and variety classes agricultural product establish growth cycle |
US20210289692A1 (en) * | 2017-12-05 | 2021-09-23 | Jiangsu University | Multi-Scale Habitat Information-Based Method and Device For Detecting and Controlling Water and Fertilizer For Crops In Seedling Stage |
-
2020
- 2020-11-16 CN CN202011275810.XA patent/CN112488467A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106707767A (en) * | 2017-03-13 | 2017-05-24 | 山东农业大学 | System and method for integrally and intelligently controlling water and fertilizer in field based on multi-source information fusion |
US20210289692A1 (en) * | 2017-12-05 | 2021-09-23 | Jiangsu University | Multi-Scale Habitat Information-Based Method and Device For Detecting and Controlling Water and Fertilizer For Crops In Seedling Stage |
CN108762079A (en) * | 2018-06-04 | 2018-11-06 | 河海大学常州校区 | The traversing process control system and method for cutter suction dredger based on deeply study |
CN109511349A (en) * | 2018-12-20 | 2019-03-26 | 山东锋士信息技术有限公司 | A kind of intelligent accurate fertigation device and control method for realizing that nutrient matches in real time |
CN109673480A (en) * | 2019-02-22 | 2019-04-26 | 福州阿里他巴信息科技有限公司 | Intelligent water and fertilizer integrated sprinkling irrigation system and control method |
CN110246051A (en) * | 2019-05-23 | 2019-09-17 | 成都信息工程大学 | The method that crop cycle model and variety classes agricultural product establish growth cycle |
CN110187688A (en) * | 2019-06-14 | 2019-08-30 | 青岛农业大学 | Industrialized agriculture intelligence liquid manure integral control system and control method |
CN110214506A (en) * | 2019-06-26 | 2019-09-10 | 北京农业智能装备技术研究中心 | Liquid manure management-control method and system |
Non-Patent Citations (2)
Title |
---|
徐凡;赵倩;贾冬冬;聂铭君;田野;段丹丹;郭文忠;: "循环式岩棉栽培水肥一体化装备及控制系统研发", 农业工程, no. 04, pages 45 - 49 * |
徐志福;李笑;徐刚;牟豪杰;汪一婷;吕永平;: "植物快繁数字化模型与智能化管理平台的构建", 浙江农业学报, no. 06, pages 461 - 466 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114662609A (en) * | 2022-04-01 | 2022-06-24 | 电子科技大学成都学院 | Intelligent greenhouse farm management method and system |
CN117437080A (en) * | 2023-12-20 | 2024-01-23 | 潍坊汇金海物联网技术有限公司 | Farmland soil water and fertilizer nutrition supplementing system and method |
CN117437080B (en) * | 2023-12-20 | 2024-03-19 | 潍坊汇金海物联网技术有限公司 | Farmland soil water and fertilizer nutrition supplementing system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109874477B (en) | Agricultural park fertilizer applicator hosting method and system | |
CN109673480A (en) | Intelligent water and fertilizer integrated sprinkling irrigation system and control method | |
CN112450056A (en) | Water, fertilizer and pesticide integrated intelligent irrigation system based on machine learning algorithm | |
CN111507857A (en) | Digital agricultural planting system and method based on Internet of things technology | |
CN112488467A (en) | Water planting crop fertilizer injection unit based on multiscale habitat information | |
CN111418332A (en) | Intelligent agriculture real-time monitoring management system based on big data | |
CN112882517B (en) | Intelligent agricultural planting environment monitoring method and cloud monitoring platform based on big data and Internet of things | |
CN106934724A (en) | Green vegetables growing way analysis system based on Internet of Things | |
CN205883318U (en) | Agricultural thing networking remote monitoring and control system | |
CN107830891B (en) | Data processing method based on aquaculture water pH value multi-parameter data acquisition device | |
CN113642729A (en) | Intelligent biogas slurry application method and system based on machine learning and storage medium | |
CN112001242A (en) | Intelligent gardening management method and device | |
CN114158336A (en) | Water and fertilizer integrated optimization control system and method | |
CN109006652B (en) | Tilapia seed production and seedling raising system and method based on Internet of things | |
CN112470907A (en) | Water planting fertilizer distributor | |
CN114895731A (en) | Breeding monitoring system and method based on digital twins | |
CN108985664A (en) | A kind of intelligence farm field monitoring management system and method | |
CN105869061A (en) | Method for analyzing plant growth state through computer | |
CN108703049A (en) | A kind of full growth cycle manuring irrigation system of crop | |
US20230309464A1 (en) | Method and apparatus for automated crop recipe optimization | |
CN208581642U (en) | A kind of full growth cycle manuring irrigation system device of crop | |
CN107633221A (en) | A kind of Fish behavior quantization system based on machine vision | |
CN115690609A (en) | Method for monitoring growth vigor of crops in growth period | |
CN111524024A (en) | Intelligent water and fertilizer irrigation system based on big data and analysis method | |
CN115088566A (en) | Rice cultivation system and cultivation method thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |