CN116910490B - Method, device, equipment and medium for adjusting environment of agricultural greenhouse - Google Patents
Method, device, equipment and medium for adjusting environment of agricultural greenhouse Download PDFInfo
- Publication number
- CN116910490B CN116910490B CN202311153863.8A CN202311153863A CN116910490B CN 116910490 B CN116910490 B CN 116910490B CN 202311153863 A CN202311153863 A CN 202311153863A CN 116910490 B CN116910490 B CN 116910490B
- Authority
- CN
- China
- Prior art keywords
- sample
- target
- agricultural greenhouse
- crop
- feature
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000007613 environmental effect Effects 0.000 claims abstract description 113
- 230000008859 change Effects 0.000 claims abstract description 109
- 230000012010 growth Effects 0.000 claims abstract description 105
- 230000002787 reinforcement Effects 0.000 claims abstract description 83
- 238000001514 detection method Methods 0.000 claims description 66
- 235000013399 edible fruits Nutrition 0.000 claims description 46
- 230000004927 fusion Effects 0.000 claims description 43
- 238000012549 training Methods 0.000 claims description 34
- 238000000605 extraction Methods 0.000 claims description 33
- 238000003062 neural network model Methods 0.000 claims description 24
- 230000008569 process Effects 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 8
- 238000011156 evaluation Methods 0.000 claims description 8
- 230000003698 anagen phase Effects 0.000 claims description 4
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 230000001276 controlling effect Effects 0.000 description 15
- 230000003750 conditioning effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 5
- 230000001105 regulatory effect Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000013507 mapping Methods 0.000 description 3
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 229910002092 carbon dioxide Inorganic materials 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 230000001143 conditioned effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000035784 germination Effects 0.000 description 1
- 238000003973 irrigation Methods 0.000 description 1
- 230000002262 irrigation Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000017260 vegetative to reproductive phase transition of meristem Effects 0.000 description 1
Abstract
The invention provides a method, a device, equipment and a medium for adjusting the environment of an agricultural greenhouse, and relates to the technical field of artificial intelligence, wherein the method comprises the following steps: acquiring a crop growth stage in a target agricultural greenhouse; when the current time step is finished, inputting a time period corresponding to the next time step, a crop growth stage and environmental change information in the target agricultural greenhouse into the trained target reinforcement learning model, acquiring equipment regulation information of the next time step output by the target reinforcement learning model, and controlling environmental equipment based on the equipment regulation information in the next time step; the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target rewards. The intelligent automatic environment adjusting device can realize intelligent automatic environment adjustment in the agricultural greenhouse.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an environment adjusting method, device, equipment and medium for an agricultural greenhouse.
Background
The environment in the greenhouse has great influence on the growth of crops, and in the prior art, various sensors can be arranged in the greenhouse to collect environmental parameters such as temperature, humidity and the like in the greenhouse, and planting personnel can adjust environmental adjusting equipment in the greenhouse according to the environmental parameters. But the planting personnel adjusts the environmental conditioning equipment in the greenhouse according to the environmental parameters, and the intelligent automatic conditioning cannot be realized because the environmental conditioning equipment is adjusted on the basis of manual experience.
Disclosure of Invention
The invention provides an environment adjusting method, device, equipment and medium for an agricultural greenhouse, which are used for solving the defect that the environment adjusting equipment in the agricultural greenhouse is adjusted by relying on manual experience in the prior art and realizing automatic adjustment of the environment in the agricultural greenhouse.
The invention provides an environment adjusting method for an agricultural greenhouse, which comprises the following steps:
acquiring a crop growth stage in a target agricultural greenhouse;
when the current time step is finished, inputting a time period corresponding to the next time step, the crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters and state information of environmental equipment of the target agricultural greenhouse in the current time step and the first N time steps of the current time step, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1;
controlling the environmental device based on the device adjustment information at the next time step;
the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward values corresponding to the sample data are determined based on environmental parameters in a sample agricultural greenhouse after the sample data are input into the target reinforcement learning model and sample adjustment information output by the target reinforcement learning model is used for controlling sample environmental equipment.
According to the method for adjusting the environment of the agricultural greenhouse, provided by the invention, the training process of the target reinforcement learning model comprises the following steps:
selecting target sample data from a plurality of groups of sample data;
inputting the target sample data into the target reinforcement learning model, and acquiring sample adjustment information corresponding to the target sample data output by the target reinforcement learning model;
adjusting sample environment equipment in the sample agricultural greenhouse based on the sample adjustment information to obtain adjusted environment parameters;
determining a reference environmental parameter based on the sample period and the sample crop growth phase in the target sample data;
determining a first prize value based on the adjusted environmental parameter and the reference environmental parameter;
determining the target prize value based on the first prize value;
updating the target reinforcement learning model based on the target reward value.
According to the method for adjusting the environment of the agricultural greenhouse provided by the invention, the determining the target rewards value based on the first rewards value comprises the following steps:
acquiring a current training round value;
when the training round number reaches a preset threshold value M+1, acquiring growth state data of crops in the sample agricultural greenhouse in the current time step and the first M time steps of the current time step;
Acquiring an evaluation score aiming at the growth state data, determining a weight value based on the evaluation score, weighting the first reward value based on the weight value to obtain the target reward value, and carrying out zero clearing on the training round value;
and when the training round number value does not reach the preset threshold value, taking the first rewarding value as the target rewarding value.
According to the method for adjusting the environment of the agricultural greenhouse, which is provided by the invention, the environmental equipment is controlled based on the equipment adjusting information in the next time step, and the method comprises the following steps:
acquiring size difference information between the target agricultural greenhouse and the sample agricultural greenhouse and configuration difference information of environment adjusting equipment;
correcting the equipment adjusting information based on the size difference information and the environment adjusting equipment configuration difference information to obtain corrected adjusting information;
controlling the environmental device based on the revised adjustment information.
According to the method for adjusting the environment of the agricultural greenhouse provided by the invention, the method for acquiring the crop growth stage in the target agricultural greenhouse comprises the following steps:
controlling an image acquisition device to acquire images of planting areas in the target agricultural greenhouse to obtain crop images;
Inputting a crop image sequence into a trained neural network model, and acquiring the crop growth stage output by the neural network model, wherein the crop image sequence comprises a plurality of crop images which are ordered in time sequence.
According to the method for adjusting the environment of the agricultural greenhouse, the neural network model comprises a target detection module, a first feature extraction module, a second feature extraction module, a third feature extraction module, a feature fusion module and a prediction module;
the step of inputting the crop image sequence into a trained neural network model, and obtaining the crop growth stage output by the neural network model comprises the following steps:
respectively inputting each crop image in the crop image sequence into the target detection module to obtain a target detection result in each crop image output by the target detection module, wherein the target detection result comprises a leaf area detection result, a flower area detection result and a fruit area detection result;
inputting the leaf area detection result corresponding to each crop image into the first feature extraction module, and obtaining leaf area change features corresponding to the crop image sequence;
Inputting the flower region detection result corresponding to each crop image into the second feature extraction module, and obtaining the flower region change feature corresponding to the crop image sequence;
inputting the fruit area detection result corresponding to each crop image into the third feature extraction module, and obtaining the fruit area change feature corresponding to the crop image sequence;
fusing the leaf area change characteristics, the flower area change characteristics and the fruit area change characteristics through the characteristic fusion module to obtain fusion characteristics;
and inputting the fusion characteristics into the prediction module, and obtaining the crop growth stage output by the prediction module.
According to the method for adjusting the environment of the agricultural greenhouse, provided by the invention, the training process of the neural network model comprises the following steps:
respectively inputting each sample crop image in a sample crop image sequence into the target detection module, and acquiring a sample leaf area detection result, a sample flower area detection result and a sample fruit area detection result which are output by the target detection module;
inputting the sample leaf area detection result corresponding to each sample crop image into the first feature extraction module, inputting the sample flower area detection result corresponding to each sample crop image into the second feature extraction module, and inputting the sample fruit area detection result corresponding to each sample crop image into the third feature extraction module to obtain sample leaf area change features, sample flower area change features and sample fruit area change features;
Fusing the sample leaf area change characteristics, the sample flower area change characteristics and the sample fruit area change characteristics through the characteristic fusion module to obtain first sample fusion characteristics;
fusing the sample leaf area change feature, the sample flower area change feature and a first preset feature through the feature fusion module to obtain a second sample fusion feature, wherein the first preset feature and the sample fruit area change feature are the same in size, and each feature value in the first preset feature is 1;
fusing the sample leaf area change feature, the sample fruit area change feature and a second preset feature through the feature fusion module to obtain a third sample fusion feature, wherein the second preset feature and the sample flower area change feature are the same in size, and each feature value in the second preset feature is 1;
inputting the first sample fusion characteristic into the prediction module to obtain a first sample crop growth stage, inputting the second sample fusion characteristic into the prediction module to obtain a second sample crop growth stage, and inputting the third sample fusion characteristic into the prediction module to obtain a third sample crop growth stage;
Obtaining a first loss according to the crop growth stage labels corresponding to the first sample crop growth stage and the sample crop image sequence, obtaining a second loss according to the second sample crop growth stage and the crop growth stage labels, and obtaining a third loss according to the third sample crop growth stage and the crop growth stage labels;
obtaining training loss according to the first loss, the second loss and the third loss, and updating parameters of a neural network model according to the training loss.
The invention also provides an environment adjusting device for the agricultural greenhouse, which comprises:
the information acquisition module is used for acquiring the growth stage of crops in the target agricultural greenhouse;
the information prediction module is used for inputting a time period corresponding to the next time step, the crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model when the current time step is finished, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters and state information of environmental equipment of the target agricultural greenhouse in the current time step and the previous N time steps of the current time step, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1;
A device control module for controlling the environmental device based on the device adjustment information at the next time step;
the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward values corresponding to the sample data are determined based on environmental parameters in a sample agricultural greenhouse after the sample data are input into the target reinforcement learning model and sample adjustment information output by the target reinforcement learning model is used for controlling sample environmental equipment.
The invention also provides electronic equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the method for regulating the environment of the agricultural greenhouse when executing the computer program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of adjusting an agricultural greenhouse environment as described in any of the above.
According to the method, the device, the equipment and the medium for adjusting the environment of the agricultural greenhouse, provided by the invention, the crop generation stage in the target agricultural greenhouse, the time period corresponding to the next time step and the environment change information in the target agricultural greenhouse are input into the trained target reinforcement learning model, the trained target reinforcement learning model can output equipment adjusting information based on the input data, and environmental equipment in the target agricultural greenhouse is controlled based on the equipment adjusting information in the next time step, so that intelligent automatic adjustment of the environment in the agricultural greenhouse is realized.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of an environment adjusting method for an agricultural greenhouse;
FIG. 2 is a schematic structural view of the environmental conditioning device for agricultural greenhouse provided by the invention;
Fig. 3 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The method for adjusting the environment of the agricultural greenhouse provided by the invention is described below with reference to fig. 1, and as shown in fig. 1, the method comprises the following steps:
s100, acquiring a crop growth stage in the target agricultural greenhouse.
The requirements of the crops on the growth environment are different in different crop growth stages, such as germination period, flowering period, fruiting period and the like. The method for obtaining the crop growth stage in the target agricultural greenhouse comprises the following steps:
controlling an image acquisition device to acquire images of planting areas in a target agricultural greenhouse to obtain crop images;
and inputting a crop image sequence into the trained neural network model, and acquiring a crop growth stage output by the neural network model, wherein the crop image sequence comprises a plurality of crop images sequenced in time sequence.
The image acquisition equipment can be arranged in the target agricultural greenhouse to acquire images of the planting areas, and crops are planted in the planting areas, so that the images of the planting areas are acquired to obtain images reflecting the growth conditions of the crops. The crop image sequence formed by the crop images obtained by collecting the images of the planting areas can reflect the growth change condition of crops, and the growth change condition of the crops can reflect the growth stage of the crops. The method provided by the invention inputs the crop image into the trained neural network model, extracts the characteristics related to the crop growth stage from the crop image sequence by utilizing the prediction capability of the neural network model, and outputs the crop growth stage.
Specifically, the neural network model for outputting the crop growth stage comprises a target detection module, a first feature extraction module, a second feature extraction module, a third feature extraction module, a feature fusion module and a prediction module. Inputting the crop image sequence into a trained neural network model, and acquiring a crop growth stage output by the neural network model, wherein the crop growth stage comprises the following steps:
respectively inputting each crop image in the crop image sequence into a target detection module to obtain a target detection result in each crop image output by the target detection module, wherein the target detection result comprises a leaf area detection result, a flower area detection result and a fruit area detection result;
Inputting a leaf area detection result corresponding to each crop image into a first feature extraction module, and acquiring leaf area change features corresponding to a crop image sequence;
inputting a flower region detection result corresponding to each crop image into a second feature extraction module, and obtaining a flower region change feature corresponding to a crop image sequence;
inputting a fruit area detection result corresponding to each crop image into a third feature extraction module, and obtaining a fruit area change feature corresponding to a crop image sequence;
fusing the leaf area change characteristics, the flower area change characteristics and the fruit area change characteristics to obtain fusion characteristics;
and inputting the fusion characteristics into a prediction module, and obtaining the crop growth stage output by the prediction module.
In the method provided by the invention, firstly, a target detection module is utilized to carry out target detection on each crop image in a crop image sequence, and the leaf, flower and fruit areas in the crop image sequence are detected. According to the leaf area detection result in each crop image in the crop image sequence, leaf area change characteristics reflecting the leaf change condition of the crops in the period corresponding to the crop image sequence can be obtained, and by adopting a similar method, flower area change characteristics reflecting the flower change condition of the crops in the period corresponding to the crop image sequence and fruit area change characteristics reflecting the fruit change condition of the crops in the period corresponding to the crop image sequence can be obtained.
Specifically, inputting a leaf area detection result corresponding to each crop image into a first feature extraction module, and acquiring leaf area change features corresponding to a crop image sequence, wherein the leaf area change features comprise:
by the first feature extraction module, the following steps are performed:
updating the pixel value of the pixel point in the blade area in the crop image to be the pixel average value of the pixel point in the blade area, and updating the pixel value of the pixel point outside the blade area in the crop image to be 0 to obtain a blade characteristic diagram;
and respectively acquiring fluctuation parameters of pixel values at each same pixel coordinate in each blade characteristic diagram to obtain the blade region change characteristics.
For each crop image, updating the pixel value of the pixel point in the blade area in the crop image to be the average value of the pixel values in the blade area, and updating the pixel value of the pixel point which is not the blade area to be 0, so that a new image can be obtained as a blade characteristic diagram. The size of the blade pixel change matrix is consistent with the size of the blade feature map, for example, the width of the blade feature map is w, the height of the blade feature map is h, then the blade pixel change matrix is a matrix of h rows and w columns, the value at the position of coordinates (x, y) in the blade pixel change matrix is a fluctuation parameter of a numerical sequence formed by pixel values at the position of coordinates (x, y) in each blade feature map, the fluctuation parameter reflects the fluctuation degree of the numerical sequence, and standard deviation, variance and the like can be selected by the fluctuation parameter.
Inputting the detection result of the flower region corresponding to each crop image into a second feature extraction module, and obtaining the change feature of the flower region corresponding to the crop image sequence, wherein the detection result comprises the following steps:
by the second feature extraction module, the following steps are performed:
updating the pixel value of the pixel point in the flower area in the crop image to be the pixel average value of the pixel point in the flower area, and updating the pixel value of the pixel point outside the flower area in the crop image to be 0 to obtain a flower characteristic diagram;
and respectively acquiring fluctuation parameters of pixel values at each same pixel coordinate in each flower characteristic diagram to obtain the change characteristics of the flower region.
Inputting the fruit area detection result corresponding to each crop image into a third feature extraction module, and acquiring the fruit area change feature corresponding to the crop image sequence, wherein the fruit area change feature comprises the following steps:
through the third feature extraction module, the following steps are performed:
updating the pixel value of the pixel point in the fruit area in the crop image to be the pixel mean value of the pixel point in the fruit area, and updating the pixel value of the pixel point outside the fruit area in the crop image to be 0 to obtain a fruit feature map;
and respectively acquiring fluctuation parameters of pixel values at each same pixel coordinate in each fruit feature map to obtain fruit region change features.
The process of obtaining the change characteristics of the flower area and the change characteristics of the fruit area is similar to the process of obtaining the change characteristics of the leaf area, and will not be described in detail here.
And fusing the leaf area change characteristics, the flower area change characteristics and the fruit area change characteristics through a characteristic fusion module, and inputting the fusion characteristics into a prediction module after obtaining fusion characteristics so as to obtain the crop growth stage output by the prediction module.
Further, the training process of the neural network model includes:
respectively inputting each sample crop image in the sample crop image sequence into a target detection module, and obtaining a sample leaf area detection result, a sample flower area detection result and a sample fruit area detection result which are output by the target detection module;
inputting a sample blade area detection result corresponding to each sample crop image into a first feature extraction module to obtain a sample blade area change feature;
inputting a sample flower region detection result corresponding to each sample crop image into a second feature extraction module to obtain a sample flower region change feature;
inputting a sample fruit area detection result corresponding to each sample crop image into a third feature extraction module to obtain a sample fruit area change feature;
The method comprises the steps that a feature fusion module fuses a sample leaf area change feature, a sample flower area change feature and a sample fruit area change feature to obtain a first sample fusion feature;
inputting the first sample fusion characteristics into a prediction module to obtain a first sample crop growth stage;
fusing the sample leaf area change feature, the sample flower area change feature and the first preset feature through a feature fusion module to obtain a second sample fusion feature, wherein the first preset feature and the sample fruit area change feature are the same in size, and each feature value in the first preset feature is 1;
inputting the second sample fusion characteristics into a prediction module to obtain a second sample crop growth stage;
fusing the sample leaf area change feature, the sample fruit area change feature and the second preset feature through a feature fusion module to obtain a third sample fusion feature, wherein the second preset feature and the sample flower area change feature are the same in size, and each feature value in the second preset feature is 1;
inputting the third sample fusion characteristics into a prediction module to obtain a third sample crop growth stage;
obtaining a first loss according to the crop growth stage labels corresponding to the first sample crop growth stage and the sample crop image sequence, obtaining a second loss according to the crop growth stage labels of the second sample crop and the crop growth stage labels, and obtaining a third loss according to the crop growth stage labels of the third sample crop and the crop growth stage labels;
Obtaining training loss according to the first loss, the second loss and the third loss, and updating parameters of the neural network model according to the training loss.
In order to improve the performance of each module in the neural network model, the method provided by the invention not only obtains the loss for updating the model parameters according to the prediction result and the label of the model in the traditional supervised learning, but also reflects the information of the growth stage of the crop based on the change of the leaves, flowers and fruits of the crop, and the characteristic that the change information of the flowers and the fruits has higher reflecting degree on the special growth stage of the crop. If the detection result of the target detection module is more accurate, the fusion module can retain the information which is more relevant to the crop growth stage in the input change characteristics, and the second sample crop growth stage and the third sample crop growth stage are more accurate.
Meanwhile, in order to meet the requirement of the fusion module on the size of the input data, a first preset feature and a second preset feature are designed to meet the data size requirement, and the values of all elements in the first preset feature and the second preset feature are set to be 1 so as to avoid influence on the result.
As shown in FIG. 1, the method for adjusting the environment of the agricultural greenhouse provided by the invention further comprises the following steps:
s200, inputting a time period corresponding to the next time step, a crop growth stage and environment dialectic information in the target agricultural greenhouse into the trained target reinforcement learning model when the current time step is finished, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environment change information comprises environment parameters and state information of environment equipment of the target agricultural greenhouse in the current time step and the previous N time steps of the current time step, the environment equipment is used for changing the environment parameters in the target agricultural greenhouse, and N is a positive integer greater than 1.
S300, controlling the environmental equipment based on the equipment adjusting information in the next time step.
The target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, the target reward values corresponding to the sample data are determined based on environmental parameters in a sample agricultural greenhouse after the sample data are input into the target reinforcement learning model and the sample environment equipment is controlled according to sample regulation information output by the target reinforcement learning model. The sample agricultural greenhouse and the target agricultural greenhouse can be the same agricultural greenhouse with the same size and the same environmental equipment configuration, and can also be different agricultural greenhouses with the size and the different environmental equipment configuration of the target agricultural greenhouse.
One time step may be one hour, two hours, etc., and at the end of the current time step, it is predicted by the target reinforcement learning how the adjusting device should be controlled for the next time step. The time period corresponding to the next time step refers to the time period of the next time step in one day, in the invention, one day can be divided into a plurality of time periods according to the growth habit of crops, and the time period corresponding to the next time step can be morning, noon, night and the like, and of course, the one day can be divided more finely. And the environmental change information of the current time step and a plurality of previous time steps is also used as the input of the target reinforcement learning model, so that the device regulation information of the next time step is predicted, and the input of the device regulation information which is more beneficial to the growth of crops is facilitated. The environmental device is a device for changing environmental parameters within an agricultural greenhouse, such as an irrigation device, a carbon dioxide concentration adjustment device, and the like.
The target reinforcement learning model is trained in a reinforcement learning mode, the reinforcement learning model is an agent in the reinforcement learning process, actions are executed according to the output result, and the agent is adjusted according to the rewarding value of the actions. The specific training process of the target reinforcement learning model in the present invention is described below. The training process of the target reinforcement learning model comprises the following steps:
Selecting target sample data from the plurality of groups of sample data;
inputting the target sample data into a target reinforcement learning model, and obtaining sample adjustment information corresponding to the target sample data output by the target reinforcement learning model;
sample environmental equipment in the sample agricultural greenhouse is regulated based on the sample regulation information, and regulated environmental parameters are obtained;
determining a reference environmental parameter based on a sample period and a sample crop growth phase in the target sample data;
determining a first prize value based on the adjusted environmental parameter and the reference environmental parameter;
determining a target prize value based on the first prize value;
the target reinforcement learning model is updated based on the target reward value.
According to the method provided by the invention, the target reinforcement learning model is adopted to output sample regulation information in the sample agricultural greenhouse, sample environmental equipment in the sample agricultural greenhouse is regulated based on the sample regulation information, and the target rewarding value is determined according to the regulated environmental parameters, so that the training of the target reinforcement learning model is realized.
Specifically, the reference environmental parameters can be determined by a plant manual of the crop, empirical knowledge can be recorded in the plant manual of the crop, the proper environmental parameter range corresponding to what growth stage of the crop is included, and the quality of sample adjustment information output by the target reinforcement learning model can be evaluated according to the adjusted environmental parameters and the reference environmental parameters to obtain a reward value for updating the target reinforcement learning model. However, in general, parameters appropriate for crop growth in empirical knowledge tend to be a range, and a number of factors affect crop growth: the temperature, the humidity, the soil nutrients and the like are mutually influenced, different growth states can be caused by each link parameter in a parameter range suitable for crop growth, and only the difference between the reference environment parameter and the adjusted environment parameter is used for determining a target rewarding value to update the target reinforcement learning model, so that the target reinforcement learning model can only output values in the range, and the target reinforcement learning model cannot output more accurate regulation information beneficial to crop growth. According to the method provided by the invention, in order to improve the performance of the target reinforcement learning model, the target reinforcement learning model can output the regulation information which is more beneficial to the growth of crops, and the target rewarding value is obtained based on the actual growth state data of the crops in the training process. Specifically, determining a target prize value based on the first prize value includes:
Acquiring a current training round value;
when the training round number reaches a preset threshold M+1, acquiring growth state data of crops in the sample agricultural greenhouse in the current time step and the first M time steps of the current time step;
acquiring an evaluation score aiming at growth state data, determining a weight value based on the evaluation score, carrying out weighting treatment on the first rewarding value based on the weight value to obtain a target rewarding value, and carrying out zero clearing treatment on the training round value;
when the training round number value does not reach the preset threshold value, the first rewarding value is used as a target rewarding value, and 1 is added to the training round number value.
The influence of the growth environment on the growth state of crops is not immediately shown, and when the target reinforcement learning model is trained, the method provided by the invention calculates the target rewarding value based on the growth state of the crops at intervals (M+1 time steps) and feeds back the target reinforcement learning model, so that the target reinforcement learning model can learn the equipment regulation information of the environmental parameters which are in fact beneficial to the growth state of the crops besides learning the equipment regulation information of the environmental parameters which are in theory suitable for the growth of the crops.
Further, controlling the environmental device based on the device adjustment information at a next time step includes:
Acquiring size difference information between a target agricultural greenhouse and a sample agricultural greenhouse and configuration difference information of environment adjusting equipment;
correcting the equipment adjusting information based on the size difference information and the environment equipment configuration difference information to obtain corrected adjusting information;
the environmental device is controlled based on the correction adjustment information.
As described above, the sizes of the target agricultural greenhouse and the sample agricultural greenhouse and the configuration conditions of the environmental conditioning equipment can be the same or different, and since the target reinforcement learning is based on the training of the sample agricultural greenhouse, the method provided by the invention corrects the equipment conditioning information output by the target reinforcement learning model based on the size difference information between the target agricultural greenhouse and the sample agricultural greenhouse and the configuration conditions of the environmental conditioning equipment when the sizes of the target agricultural greenhouse and the sample agricultural greenhouse and the configuration conditions of the environmental conditioning equipment are different, so as to improve the adaptability of the target reinforcement learning model, and controls the environmental equipment in the target agricultural greenhouse based on the corrected and conditioned information.
Specifically, the size difference information reflects the difference of the sizes of the target agricultural greenhouse and the sample agricultural greenhouse, and the environmental equipment configuration difference information reflects the difference of equipment configured in the target agricultural greenhouse and the sample agricultural greenhouse for adjusting environmental parameters, thereby being beneficial to the number, the model and the like of the environmental equipment. The device adjustment information is corrected based on the size difference information and the environment device configuration difference information to obtain corrected adjustment information, which may be determined according to a pre-established adjustment information mapping relationship between the target agricultural greenhouse and the sample agricultural greenhouse, that is, a pre-established mapping relationship between the adjustment information of the environment devices corresponding to the target agricultural greenhouse and the sample agricultural greenhouse respectively when the environmental parameter variation amounts are the same, where the mapping relationship may be linear or nonlinear.
The apparatus for adjusting the environment of an agricultural greenhouse provided by the invention is described below, and the apparatus for adjusting the environment of an agricultural greenhouse described below and the method for adjusting the environment of an agricultural greenhouse described above can be referred to correspondingly with each other. As shown in fig. 2, the apparatus includes:
an information acquisition module 210 for acquiring a crop growth stage in a target agricultural greenhouse;
the information prediction module 220 is configured to input, when the current time step is over, a time period corresponding to the next time step, a crop growth stage, and environmental change information in the target agricultural greenhouse into the trained target reinforcement learning model, and obtain device adjustment information of the next time step output by the target reinforcement learning model, where the environmental change information includes environmental parameters of the target agricultural greenhouse in the current time step and the first N time steps of the current time step and state information of environmental devices, and the environmental devices are used to change the environmental parameters in the target agricultural greenhouse, where N is a positive integer greater than 1;
a device control module 230 for controlling the environmental device based on the device adjustment information at a next time step;
the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward value corresponding to the sample data is determined based on environmental parameters in a sample agricultural greenhouse after the sample data is input into the target reinforcement learning model and sample regulation information output by the target reinforcement learning model is used for controlling sample environment equipment.
Fig. 3 illustrates a physical schematic diagram of an electronic device, as shown in fig. 3, where the electronic device may include: processor 310, communication interface (Communications Interface) 320, memory 330 and communication bus 340, wherein processor 310, communication interface 320, memory 330 accomplish communication with each other through communication bus 340. The processor 310 may invoke logic instructions in the memory 330 to perform an agricultural greenhouse environment adjustment method comprising: acquiring a crop growth stage in a target agricultural greenhouse; when the current time step is finished, inputting a time period corresponding to the next time step, a crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters of the target agricultural greenhouse in the current time step and the previous N time steps of the current time step and state information of environmental equipment, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1; controlling the environmental device based on the device adjustment information at a next time step; the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward value corresponding to the sample data is determined based on environmental parameters in a sample agricultural greenhouse after the sample data is input into the target reinforcement learning model and sample regulation information output by the target reinforcement learning model is used for controlling sample environment equipment.
Further, the logic instructions in the memory 330 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In still another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the method for adjusting an environment of an agricultural greenhouse provided by the above methods, the method comprising: acquiring a crop growth stage in a target agricultural greenhouse; when the current time step is finished, inputting a time period corresponding to the next time step, a crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters of the target agricultural greenhouse in the current time step and the previous N time steps of the current time step and state information of environmental equipment, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1; controlling the environmental device based on the device adjustment information at a next time step; the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward value corresponding to the sample data is determined based on environmental parameters in a sample agricultural greenhouse after the sample data is input into the target reinforcement learning model and sample regulation information output by the target reinforcement learning model is used for controlling sample environment equipment.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (8)
1. An agricultural greenhouse environment adjustment method is characterized by comprising the following steps:
acquiring a crop growth stage in a target agricultural greenhouse;
when the current time step is finished, inputting a time period corresponding to the next time step, the crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters and state information of environmental equipment of the target agricultural greenhouse in the current time step and the first N time steps of the current time step, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1;
Controlling the environmental device based on the device adjustment information at the next time step;
the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward value corresponding to the sample data is determined based on environmental parameters in a sample agricultural greenhouse after the sample data is input into the target reinforcement learning model and sample adjustment information output by the target reinforcement learning model is used for controlling sample environmental equipment;
the training process of the target reinforcement learning model comprises the following steps:
selecting target sample data from a plurality of groups of sample data;
inputting the target sample data into the target reinforcement learning model, and acquiring sample adjustment information corresponding to the target sample data output by the target reinforcement learning model;
adjusting sample environment equipment in the sample agricultural greenhouse based on the sample adjustment information to obtain adjusted environment parameters;
determining a reference environmental parameter based on the sample period and the sample crop growth phase in the target sample data;
Determining a first prize value based on the adjusted environmental parameter and the reference environmental parameter;
determining the target prize value based on the first prize value;
updating the target reinforcement learning model based on the target reward value;
the determining the target prize value based on the first prize value includes:
acquiring a current training round value;
when the training round number reaches a preset threshold value M+1, acquiring growth state data of crops in the sample agricultural greenhouse in the current time step and the first M time steps of the current time step;
acquiring an evaluation score aiming at the growth state data, determining a weight value based on the evaluation score, weighting the first reward value based on the weight value to obtain the target reward value, and carrying out zero clearing on the training round value;
and when the training round number value does not reach the preset threshold value, taking the first rewarding value as the target rewarding value.
2. The agricultural greenhouse environment adjustment method according to claim 1, wherein the controlling the environmental apparatus based on the apparatus adjustment information at the next time step includes:
Acquiring size difference information between the target agricultural greenhouse and the sample agricultural greenhouse and configuration difference information of environment adjusting equipment;
correcting the equipment adjusting information based on the size difference information and the environment adjusting equipment configuration difference information to obtain corrected adjusting information;
controlling the environmental device based on the revised adjustment information.
3. The agricultural greenhouse environment adjustment method according to any one of claims 1 to 2, wherein the acquisition of the crop growth stage in the target agricultural greenhouse includes:
controlling an image acquisition device to acquire images of planting areas in the target agricultural greenhouse to obtain crop images;
inputting a crop image sequence into a trained neural network model, and acquiring the crop growth stage output by the neural network model, wherein the crop image sequence comprises a plurality of crop images which are ordered in time sequence.
4. The method for adjusting the environment of an agricultural greenhouse according to claim 3, wherein the neural network model comprises a target detection module, a first feature extraction module, a second feature extraction module, a third feature extraction module, a feature fusion module and a prediction module;
The step of inputting the crop image sequence into a trained neural network model, and obtaining the crop growth stage output by the neural network model comprises the following steps:
respectively inputting each crop image in the crop image sequence into the target detection module to obtain a target detection result in each crop image output by the target detection module, wherein the target detection result comprises a leaf area detection result, a flower area detection result and a fruit area detection result;
inputting the leaf area detection result corresponding to each crop image into the first feature extraction module, and obtaining leaf area change features corresponding to the crop image sequence;
inputting the flower region detection result corresponding to each crop image into the second feature extraction module, and obtaining the flower region change feature corresponding to the crop image sequence;
inputting the fruit area detection result corresponding to each crop image into the third feature extraction module, and obtaining the fruit area change feature corresponding to the crop image sequence;
fusing the leaf area change characteristics, the flower area change characteristics and the fruit area change characteristics through the characteristic fusion module to obtain fusion characteristics;
And inputting the fusion characteristics into the prediction module, and obtaining the crop growth stage output by the prediction module.
5. The method for adjusting the environment of an agricultural greenhouse according to claim 4, wherein the training process of the neural network model comprises:
respectively inputting each sample crop image in a sample crop image sequence into the target detection module, and acquiring a sample leaf area detection result, a sample flower area detection result and a sample fruit area detection result which are output by the target detection module;
inputting the sample leaf area detection result corresponding to each sample crop image into the first feature extraction module, inputting the sample flower area detection result corresponding to each sample crop image into the second feature extraction module, and inputting the sample fruit area detection result corresponding to each sample crop image into the third feature extraction module to obtain sample leaf area change features, sample flower area change features and sample fruit area change features;
fusing the sample leaf area change characteristics, the sample flower area change characteristics and the sample fruit area change characteristics through the characteristic fusion module to obtain first sample fusion characteristics;
Fusing the sample leaf area change feature, the sample flower area change feature and a first preset feature through the feature fusion module to obtain a second sample fusion feature, wherein the first preset feature and the sample fruit area change feature are the same in size, and each feature value in the first preset feature is 1;
fusing the sample leaf area change feature, the sample fruit area change feature and a second preset feature through the feature fusion module to obtain a third sample fusion feature, wherein the second preset feature and the sample flower area change feature are the same in size, and each feature value in the second preset feature is 1;
inputting the first sample fusion characteristic into the prediction module to obtain a first sample crop growth stage, inputting the second sample fusion characteristic into the prediction module to obtain a second sample crop growth stage, and inputting the third sample fusion characteristic into the prediction module to obtain a third sample crop growth stage;
obtaining a first loss according to the crop growth stage labels corresponding to the first sample crop growth stage and the sample crop image sequence, obtaining a second loss according to the second sample crop growth stage and the crop growth stage labels, and obtaining a third loss according to the third sample crop growth stage and the crop growth stage labels;
Obtaining training loss according to the first loss, the second loss and the third loss, and updating parameters of the neural network model according to the training loss.
6. An agricultural greenhouse environment adjusting device, comprising:
the information acquisition module is used for acquiring the growth stage of crops in the target agricultural greenhouse;
the information prediction module is used for inputting a time period corresponding to the next time step, the crop growth stage and environmental change information in the target agricultural greenhouse into a trained target reinforcement learning model when the current time step is finished, and acquiring equipment adjustment information of the next time step output by the target reinforcement learning model, wherein the environmental change information comprises environmental parameters and state information of environmental equipment of the target agricultural greenhouse in the current time step and the previous N time steps of the current time step, and the environmental equipment is used for changing the environmental parameters in the target agricultural greenhouse, wherein N is a positive integer greater than 1;
a device control module for controlling the environmental device based on the device adjustment information at the next time step;
the target reinforcement learning model is trained based on a plurality of groups of randomly generated sample data and corresponding target reward values, each group of sample data comprises a sample time period, a sample crop growth stage and sample environment change information, and the target reward value corresponding to the sample data is determined based on environmental parameters in a sample agricultural greenhouse after the sample data is input into the target reinforcement learning model and sample adjustment information output by the target reinforcement learning model is used for controlling sample environmental equipment;
The training process of the target reinforcement learning model comprises the following steps:
selecting target sample data from a plurality of groups of sample data;
inputting the target sample data into the target reinforcement learning model, and acquiring sample adjustment information corresponding to the target sample data output by the target reinforcement learning model;
adjusting sample environment equipment in the sample agricultural greenhouse based on the sample adjustment information to obtain adjusted environment parameters;
determining a reference environmental parameter based on the sample period and the sample crop growth phase in the target sample data;
determining a first prize value based on the adjusted environmental parameter and the reference environmental parameter;
determining the target prize value based on the first prize value;
updating the target reinforcement learning model based on the target reward value;
the determining the target prize value based on the first prize value includes:
acquiring a current training round value;
when the training round number reaches a preset threshold value M+1, acquiring growth state data of crops in the sample agricultural greenhouse in the current time step and the first M time steps of the current time step;
Acquiring an evaluation score aiming at the growth state data, determining a weight value based on the evaluation score, weighting the first reward value based on the weight value to obtain the target reward value, and carrying out zero clearing on the training round value;
and when the training round number value does not reach the preset threshold value, taking the first rewarding value as the target rewarding value.
7. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of adjusting the environment of an agricultural greenhouse according to any one of claims 1 to 5 when executing the computer program.
8. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the method of adjusting an agricultural greenhouse environment of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311153863.8A CN116910490B (en) | 2023-09-08 | Method, device, equipment and medium for adjusting environment of agricultural greenhouse |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311153863.8A CN116910490B (en) | 2023-09-08 | Method, device, equipment and medium for adjusting environment of agricultural greenhouse |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116910490A CN116910490A (en) | 2023-10-20 |
| CN116910490B true CN116910490B (en) | 2023-12-15 |
Family
ID=
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107390753A (en) * | 2017-08-29 | 2017-11-24 | 贵州省岚林阳环保能源科技有限责任公司 | Intelligent plant growth environment regulating device and method based on Internet of Things cloud platform |
| CN109087212A (en) * | 2018-10-10 | 2018-12-25 | 贵州网安科技发展有限公司 | Agricultural production greenhouse regulation device based on big data platform |
| CN112800665A (en) * | 2021-01-13 | 2021-05-14 | 电凯(常州)新能源技术研究院有限公司 | Plant growth self-feedback learning cultivation method based on artificial intelligence |
| CN113221723A (en) * | 2021-05-08 | 2021-08-06 | 余治梅 | Traceable self-feedback learning urban plant factory |
| CN113485491A (en) * | 2021-05-20 | 2021-10-08 | 西安微媒软件有限公司 | Greenhouse intelligent detection control model and device |
| CN115292753A (en) * | 2022-07-01 | 2022-11-04 | 电子科技大学中山学院 | Agricultural greenhouse data tracing and management method based on block chain |
| CN115686110A (en) * | 2023-01-07 | 2023-02-03 | 广州市农业科学研究院 | Greenhouse intelligent control method, system, monitoring device, equipment and medium |
| CN115717758A (en) * | 2022-11-04 | 2023-02-28 | 北京邮电大学 | Indoor space temperature and humidity regulation and control method and system |
| KR102520715B1 (en) * | 2021-11-17 | 2023-04-10 | 연세대학교 산학협력단 | Method and device for automatic heating and cooling control using indoor and outdoor environment change prediction |
| CN116448117A (en) * | 2023-04-18 | 2023-07-18 | 安徽大学 | Path planning method integrating deep neural network and reinforcement learning method |
| CN116649160A (en) * | 2023-08-01 | 2023-08-29 | 南京康之春生物科技有限公司 | Edible fungus strain production monitoring system and monitoring method |
| CN116703074A (en) * | 2023-05-29 | 2023-09-05 | 中国南方电网有限责任公司 | Feedback enhancement-based intelligent power scheduling method, device, equipment and medium |
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107390753A (en) * | 2017-08-29 | 2017-11-24 | 贵州省岚林阳环保能源科技有限责任公司 | Intelligent plant growth environment regulating device and method based on Internet of Things cloud platform |
| CN109087212A (en) * | 2018-10-10 | 2018-12-25 | 贵州网安科技发展有限公司 | Agricultural production greenhouse regulation device based on big data platform |
| CN112800665A (en) * | 2021-01-13 | 2021-05-14 | 电凯(常州)新能源技术研究院有限公司 | Plant growth self-feedback learning cultivation method based on artificial intelligence |
| CN113221723A (en) * | 2021-05-08 | 2021-08-06 | 余治梅 | Traceable self-feedback learning urban plant factory |
| CN113485491A (en) * | 2021-05-20 | 2021-10-08 | 西安微媒软件有限公司 | Greenhouse intelligent detection control model and device |
| KR102520715B1 (en) * | 2021-11-17 | 2023-04-10 | 연세대학교 산학협력단 | Method and device for automatic heating and cooling control using indoor and outdoor environment change prediction |
| CN115292753A (en) * | 2022-07-01 | 2022-11-04 | 电子科技大学中山学院 | Agricultural greenhouse data tracing and management method based on block chain |
| CN115717758A (en) * | 2022-11-04 | 2023-02-28 | 北京邮电大学 | Indoor space temperature and humidity regulation and control method and system |
| CN115686110A (en) * | 2023-01-07 | 2023-02-03 | 广州市农业科学研究院 | Greenhouse intelligent control method, system, monitoring device, equipment and medium |
| CN116448117A (en) * | 2023-04-18 | 2023-07-18 | 安徽大学 | Path planning method integrating deep neural network and reinforcement learning method |
| CN116703074A (en) * | 2023-05-29 | 2023-09-05 | 中国南方电网有限责任公司 | Feedback enhancement-based intelligent power scheduling method, device, equipment and medium |
| CN116649160A (en) * | 2023-08-01 | 2023-08-29 | 南京康之春生物科技有限公司 | Edible fungus strain production monitoring system and monitoring method |
Non-Patent Citations (1)
| Title |
|---|
| 支持多平台应用的智慧农业温室大棚监控系统设计;张晨;施佺;付康为;沈乐威;郑亚平;;江苏农业科学(第12期);205-209 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6261492B2 (en) | Information processing apparatus, information processing method, and program | |
| KR20180076766A (en) | Artificial Intelligence Based Smart Farm Management System | |
| CN115204689B (en) | Intelligent agriculture management system based on image processing | |
| CN104750146A (en) | Plant growing regulating method and device | |
| WO2021007363A1 (en) | Irrigation control with deep reinforcement learning and smart scheduling | |
| CN115529987B (en) | Air port regulating and controlling method, device, equipment and storage medium for crop facility | |
| KR102265587B1 (en) | Smart Farm Control System Using Smart Farm Control History | |
| CN111095314A (en) | Yield estimation for crop plant planting | |
| JP2019170359A (en) | Plant cultivation result prediction system | |
| Ali et al. | Wireless sensor network and deep learning for prediction greenhouse environments | |
| JP2018007630A (en) | Cultivation support device | |
| Avigal et al. | Learning seed placements and automation policies for polyculture farming with companion plants | |
| CN113994868B (en) | Automatic irrigation method and system based on plant growth cycle | |
| US10064347B2 (en) | Plant cultivation system, and plant cultivation unit | |
| Tarova et al. | Analysis of taxation assessment results and development of a method for applying digital technologies in the assessment of garden agrocenoses stability | |
| CN116910490B (en) | Method, device, equipment and medium for adjusting environment of agricultural greenhouse | |
| CA3196136A1 (en) | Advanced crop manager for crops stress mitigation | |
| JP2020156431A (en) | Agricultural crop growth evaluation/work proposal method, agricultural product growth evaluation/work proposal system, and regional agricultural activation system | |
| Ahmad et al. | Speaking plant approach for automatic fertigation system in greenhouse | |
| CN116910490A (en) | Method, device, equipment and medium for adjusting environment of agricultural greenhouse | |
| Van Straten | On-line optimal control of greenhouse crop cultivation | |
| KR102440824B1 (en) | Learning system and method for optimal growth environment of tomato | |
| EP4340590B1 (en) | Greenhouse environment optimisation | |
| Kumar et al. | Monitoring and accelerating plant growth using IoT and Hydroponics | |
| KR20240070735A (en) | Apparatus for controlling irrigation of ai smart farm system for drain optimization and method for determining amount of irrigation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |