CN109828623B - Production management method and device for greenhouse crop context awareness - Google Patents

Abstract

Embodiments of the present invention provide a production management method and device for situational perception of greenhouse crops, which belong to the technical field of crop planting management. The method includes: acquiring monitoring parameters of greenhouse crops, and preprocessing the monitoring parameters; based on a deep learning model, performing situational perception on the preprocessed monitoring parameters to obtain decision information, and performing production management based on the decision information. In the method provided by the embodiment of the present invention, the monitoring parameters of the greenhouse crops are obtained, and the monitoring parameters are preprocessed. Based on the deep learning model, the preprocessed monitoring parameters are subjected to situational awareness to obtain decision-making information, and production management is carried out based on the decision-making information. Because it can sense multiple growth, physiological indicators and environmental parameters of crops and make corresponding decisions, it can change the traditional greenhouse crop management method to realize the intelligent production, monitoring and regulation of greenhouse crops, thereby improving the utilization rate of greenhouse monitoring data.

Description

Production management method and device for greenhouse crop situation perception

technical field

The embodiments of the present invention relate to the technical field of crop planting management, and in particular, to a production management method and device for contextual perception of greenhouse crops.

Background technique

At present, most control systems are built based on experience to realize greenhouse production management. To realize the automation and intelligence of greenhouse production, on the one hand, monitoring data must be supported, and on the other hand, a decision-making system must be used for technical support. In related technologies, binocular image data and growth environment data of crops are usually collected, binocular stereo vision system is used to shoot binocular images of crops, image data of crops are collected, and a meteorological moisture monitoring system is used to monitor the growth environment of crops. Data, including light, temperature, moisture, soil parameters and other indicators, combine the two data to assist decision-making in field production. Since the above process uses context-aware data for decision-making, the decision-making algorithm is relatively single, so when the types of context information increase, the algorithm will be too complicated and lengthy, which is not conducive to developers' maintenance and upgrades, and increases the overall situation-awareness system. The failure rate .

SUMMARY OF THE INVENTION

In order to solve the above problems, the embodiments of the present invention provide a production management method and device for situation awareness of greenhouse crops that overcome the above problems or at least partially solve the above problems.

According to a first aspect of the embodiments of the present invention, there is provided a production management method for contextual awareness of greenhouse crops, including:

Obtain the monitoring parameters of greenhouse crops, and preprocess the monitoring parameters. The monitoring parameters include at least growth parameters, physiological parameters and environmental parameters;

Based on the deep learning model, the preprocessed monitoring parameters are subjected to situational awareness to obtain decision information, and production management is carried out based on the decision information.

In the method provided by the embodiment of the present invention, the monitoring parameters of greenhouse crops are obtained, and the monitoring parameters are preprocessed. Based on the deep learning model, the preprocessed monitoring parameters are subjected to situational awareness to obtain decision information, and production management is carried out based on the decision information. By sensing multiple growth, physiological indicators and environmental parameters of crops and making corresponding decisions, the traditional greenhouse crop management method can be changed, so as to realize the intelligent production, monitoring and regulation of greenhouse crops, thereby improving the utilization rate of greenhouse monitoring data.

Secondly, by applying context-aware methods and concepts to greenhouse crop production, context-aware decision-making can be realized from the perspective of greenhouse crops, and the production management of greenhouse crop feedback can be realized at a lower cost. In addition, due to the modular design of the situation awareness framework, it can be combined according to different greenhouse environments and conditions, which is convenient for future maintenance and development, and then realizes multi-level integration to meet more complex production needs. Finally, because the deep learning model is used as the inference engine for situational perception, according to the structure and characteristics of the deep network, more types of situational information such as numerical values, texts, images, etc. can be processed at the same time, and the perception results are accurate and fast, which can meet the needs of the system. required timeliness and robustness.

According to a second aspect of the embodiments of the present invention, there is provided a production management device for situational awareness of greenhouse crops, including:

an acquisition module for acquiring monitoring parameters of greenhouse crops, where the monitoring parameters at least include growth parameters, physiological parameters and environmental parameters;

The preprocessing module is used to preprocess the monitoring parameters;

The context perception module is used to perform context perception on the preprocessed monitoring parameters based on the deep learning model to obtain decision information;

The production management module is used for production management based on decision information.

According to a third aspect of the embodiments of the present invention, an electronic device is provided, including:

at least one processor; and

at least one memory communicatively coupled to the processor, wherein:

The memory stores program instructions executable by the processor, and the processor invokes the program instructions to execute the production management method provided by any of the various possible implementation manners of the first aspect provided by the situational awareness of greenhouse crops.

According to a fourth aspect of the present invention, a non-transitory computer-readable storage medium is provided, the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions cause a computer to execute any of the various possible implementations of the first aspect. A possible implementation of the provided context-aware production management method for greenhouse crops.

It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory and are not limiting of embodiments of the present invention.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.

1 is a schematic flowchart of a production management method for greenhouse crop situation awareness provided by an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of a context awareness framework provided by an embodiment of the present invention;

3 is a schematic diagram of a data flow according to an embodiment of the present invention;

4 is a schematic structural diagram of a production management device for greenhouse crop situation perception provided by an embodiment of the present invention;

FIG. 5 is a block diagram of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments These are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

At present, most control systems are built based on experience to realize greenhouse production management. To realize the automation and intelligence of greenhouse production, on the one hand, monitoring data must be supported, and on the other hand, a decision-making system must be used for technical support. In related technologies, binocular image data and growth environment data of crops are usually collected, binocular stereo vision system is used to shoot binocular images of crops, image data of crops are collected, and a meteorological moisture monitoring system is used to monitor the growth environment of crops. Data, including light, temperature, moisture, soil parameters and other indicators, combine the two data to assist decision-making in field production. Since the above process uses context-aware data for decision-making, the decision-making algorithm is relatively single, so when the types of context information increase, the algorithm will be too complicated and lengthy, which is not conducive to developers' maintenance and upgrades, and increases the overall situation-awareness system. The failure rate .

In view of the above situation, the embodiment of the present invention provides a production management method for greenhouse crop situation awareness. Referring to Figure 1, the method includes:

101. Obtain monitoring parameters of the greenhouse crop, and preprocess the monitoring parameters, where the monitoring parameters at least include growth parameters, physiological parameters, and environmental parameters.

Among them, growth parameters can be obtained through cameras and/or TOF (Time of Flight) equipment, physiological parameters can be directly measured and obtained through photosynthesis meters, and environmental parameters can be obtained through data collectors arranged in the greenhouse to integrate various environmental sensors The monitoring and acquisition are not specifically limited in this embodiment of the present invention. It should be noted that, after the monitoring parameters are acquired, the acquired data may also be aggregated. Specifically, the physiological parameters can be measured using photosynthetic measuring instruments and the data can be derived for sorting. Environmental parameter information can be transmitted to the database for real-time storage and download service by using the built-in GPRS module of the data collector. Growth parameters were obtained by image processing. Finally, all kinds of acquired data information can be integrated into a CSV format data set, so that the fusion system can directly make perception decisions.

102. Based on the deep learning model, perform situational awareness on the preprocessed monitoring parameters, obtain decision information, and perform production management based on the decision information.

In the method provided by the embodiment of the present invention, the monitoring parameters of greenhouse crops are obtained, and the monitoring parameters are preprocessed. Based on the deep learning model, the preprocessed monitoring parameters are subjected to situational awareness to obtain decision information, and production management is carried out based on the decision information. By sensing multiple growth, physiological indicators and environmental parameters of crops and making corresponding decisions, the traditional greenhouse crop management method can be changed, so as to realize the intelligent production, monitoring and regulation of greenhouse crops, thereby improving the utilization rate of greenhouse monitoring data.

Secondly, by applying context-aware methods and concepts to greenhouse crop production, context-aware decision-making can be realized from the perspective of greenhouse crops, and the production management of greenhouse crop feedback can be realized at a lower cost. In addition, due to the modular design of the situation awareness framework, it can be combined according to different greenhouse environments and conditions, which is convenient for future maintenance and development, and then realizes multi-level integration to meet more complex production needs. Finally, because the deep learning model is used as the inference engine for situational perception, according to the structure and characteristics of the deep network, more types of situational information such as numerical values, texts, images, etc. can be processed at the same time, and the perception results are accurate and fast, which can meet the needs of the system. required timeliness and robustness.

Based on the content of the above embodiment, as an optional embodiment, the growth parameters include at least crop growth parameters, plant height and canopy leaf area parameters; physiological parameters include at least crop net photosynthetic rate, canopy temperature, transpiration and crop fluorescence parameters ; Environmental parameters include at least air temperature, air humidity, soil temperature, soil humidity, light intensity and CO2 concentration.

Based on the content of the foregoing embodiment, as an optional embodiment, before preprocessing the monitoring parameters, the method further includes: screening the monitoring parameters based on the type of greenhouse operation. Among them, the purpose of information screening is to select different contextual information as the input of the fusion model according to different perception targets, thereby enhancing the scalability and generalization ability of the system and avoiding decision interference caused by information redundancy.

Based on the content of the foregoing embodiment, as an optional embodiment, the method of preprocessing monitoring parameters is not specifically limited in this embodiment of the present invention, including but not limited to: removing missing value parameters in monitoring parameters, using The growth parameters, physiological parameters and environmental parameters in the monitoring parameters are matched, and the matched monitoring parameters are normalized. It should be noted that the normalization process here can be an optional step, that is, it is not necessarily beneficial to data fusion and decision-making after normalization, so whether normalization is required can be manually selected.

Based on the content of the foregoing embodiment, as an optional embodiment, the embodiment of the present invention does not specifically limit the method of performing situational awareness on the preprocessed monitoring parameters based on the deep learning model to obtain decision information, including but not limited to: Based on the deep learning model, the monitoring parameters are fused to obtain regional situation information; based on the regional situation information, the overall perception results of the greenhouse are obtained; the overall perception results of the greenhouse are matched with preset control decisions to obtain decision information.

Specifically, the deep learning model based on the Tensorflow toolkit can be trained first by using the data set; the trained model can be used as the inference engine of each middleware, and the regional context information can be obtained by fusing the data. Then, the information of each area is given the overall perception result of the greenhouse according to the principle of minority obeying the majority. Match the perception results with customized regulation decisions or services to obtain decision information. Finally, use the decision information to regulate the greenhouse crop environment or provide corresponding services.

Based on the content of the foregoing embodiment, as an optional embodiment, the decision-making information at least includes environmental regulation information, water and fertilizer management information, and agricultural operation information.

式中Y为分类类别标签、SoftMax(·)为多类别分类函数、

为原始特征x的交叉乘积转换、b为模型的偏置、W_wide为wide模型的权重向量、W_deep为deep网络的权重向量。Based on the content of the foregoing embodiment, as an optional embodiment, the deep learning model is obtained by training based on a wide-deep neural network. Among them, the wide and deep neural network can use the characteristics of joint training and joint classification of its deep network and wide network, and is responsible for accurately classifying or predicting the input environmental information, growth information and physiological information, so as to achieve the effect of data fusion, and according to the classification The category provides decision-making information on the environmental status of each area of the greenhouse, and gives suggestions for greenhouse environmental regulation. The wide and deep neural network is composed of a wide network and a deep network. The wide network is good at dealing with sparse features (such as text features), and the deep network is good at dealing with high-dimensional features (numerical features such as weather). The corresponding mathematical expression is :

where Y is the classification category label, SoftMax( ) is the multi-category classification function,

is the cross-product transformation of the original feature x, b is the bias of the model, W _wide is the weight vector of the wide model, and W _deep is the weight vector of the deep network.

Based on the production management method for greenhouse crop situation perception provided in the above embodiment, a specific application example of the method is now given in combination with the actual application scenario. Among them, the method can adjust the scale and perception category of the framework according to the needs; the context information can be manually input or crawled from the database, and an appropriate inference engine can be selected according to the characteristics of the context information to build a suitable framework.

Data acquisition: The situational information required for situational awareness needs to obtain monitoring data through the data collectors evenly distributed in the greenhouse. The monitoring equipment adopts the greenhouse cloud environment data collector, and each data collector can be installed on a tripod. Each data collector can measure air temperature, air relative humidity, soil temperature, soil humidity, CO ₂ and light intensity. The types of monitoring data can be increased or decreased according to specific production needs, and sent to the remote end at a set time interval (30min). Cloud server, the context awareness system can connect to the database interface to obtain real-time monitoring data, and administrators can also manually input context information for perception.

Data processing: Data preprocessing needs to be selected according to the characteristics of the inference engine algorithm. Before data processing, the data set is uniformly adjusted to CSV file format, which is convenient for data processing and model perception. When using the wide and deep neural network for decision-making of environmental state information in the greenhouse area, according to the comparison of experimental results, the effect of input parameters without preprocessing is better than normalization and regularization preprocessing. Therefore, no data preprocessing operation is taken when sensing environmental state information. At the same time, the administrator can manually select whether to retain the outliers in the data of the model training set according to their own production characteristics and perception needs. And according to the recognition results, it will give warnings and sensor failure prompts and other functions.

Context-aware framework construction: 6 data collection points can be selected, and 4858 pieces of data from December 23, 2017 to January 2, 2018 are used for model training and testing. Among them, there are 4009 pieces of data in the training set and 849 pieces of data in the test set. Each data was collected continuously at 30min interval and 24h. The downloaded data is classified into categories, and the training set and test set are used for the construction of the wide and deep neural network. The model structure can be selected as 7-100-50-7, and the training iteration is 2000 steps. The environment for the perception system can be the MacOS operating system, the model implementation can be based on Google's open source Tensorflow toolkit, the programming language can be Python, and the integrated development environment (IDE) can be the jupyter notebook integrated in Anaconda.

Due to the modular design of the system, developers or maintainers can add or delete framework levels according to specific needs. Each level is composed of three modules: data integration, data processing and scenario fusion. (the middleware is the class in the programming language). The middleware can encapsulate the data uploaded by the sensor, that is, convert text data into sparse data, and convert numerical data into tensor type data. Then use the inference engine of the middleware to perform situational awareness on the encapsulated data. The perception process is to classify the input features. Later, it is verified that the classification accuracy of the system (the ratio of the number of correctly classified samples to the total number of classified samples) can reach 98.00% , that is, the verification can achieve the predicted expected result.

Provide corresponding decision information according to the category, and transmit the perceived decision information to the application layer. The UI design of the application layer is implemented by TKinter toolkit programming, which can greatly facilitate the situation awareness system to call it. Finally, according to the classification results of the fusion model, the system gives the current growth stage information of crops, current environmental information, corresponding service strategies and necessary hardware control suggestions, and finally realizes real-time perception of the state of greenhouse crops, so that the relevant personnel in greenhouse production can clearly understand. Master the growth status and environmental information of greenhouse crops, and realize the effect of self-adaptive adjustment.

The context awareness framework involved in the embodiment of the present invention may refer to FIG. 2 , and the data flow of the method provided by the embodiment of the present invention may refer to FIG. 3 . Combined with the above specific application examples, the effects of the above specific application examples are as follows:

First, use the modular idea to design a situational perception framework for greenhouse crops, which can achieve compatibility with different greenhouse environments and greenhouse conditions, facilitate the addition or deletion of components of the framework, and realize complex environments with more situational information. Multi-level perception achieves more accurate decision information and has good generalization ability.

Second, the core inference engine of the context awareness framework is composed of a deep learning model. Because the deep learning model has stronger data processing capabilities and more types of processing features, it can better accommodate complex and changeable scenarios while adding or reducing framework components. Information and, in line with the requirements of modular design.

Third, because the situational awareness method can efficiently and accurately decide the environment or state of greenhouse crops, and provide corresponding services and control mechanisms, it can reduce human resources while improving management efficiency and realize the intelligence and automation of greenhouse crop production.

Fourth, apply situational awareness and deep learning to greenhouse production to realize intelligent perception of greenhouse crop environment, growth, physiology and other information. Apply SPA (Speaking Plant Approach) to the facility agricultural production management platform to achieve efficient, energy-saving and optimal management of "talking with plants". Taking the changes of plants themselves as the basis for changing their growth environment, changing the traditional environmental control and decision-making management methods.

Based on the content of the above embodiments, the embodiments of the present invention further provide a production management device for greenhouse crop situation awareness, the device being used to execute the production management method for greenhouse crop situation awareness provided in the above method embodiments. Referring to FIG. 4 , the device includes: an acquisition module 401, a preprocessing module 402, a context awareness module 403 and a production management module 404; wherein,

The acquisition module 401 is used to acquire monitoring parameters of greenhouse crops, where the monitoring parameters at least include growth parameters, physiological parameters and environmental parameters;

a preprocessing module 402, configured to preprocess the monitoring parameters;

The context perception module 403 is configured to perform context perception on the preprocessed monitoring parameters based on the deep learning model to obtain decision information;

The production management module 404 is used for production management based on the decision information.

Based on the content of the above embodiment, as an optional embodiment, the growth parameters include at least crop growth parameters, plant height and canopy leaf area parameters; physiological parameters include at least crop net photosynthetic rate, canopy temperature, transpiration and crop fluorescence parameters ; Environmental parameters include at least air temperature, air humidity, soil temperature, soil humidity, light intensity and CO2 concentration.

Based on the content of the foregoing embodiment, as an optional embodiment, the apparatus further includes:

Screening module for screening monitoring parameters based on the type of greenhouse operation.

Based on the content of the above embodiment, as an optional embodiment, the preprocessing module 402 is used to remove missing value parameters in the monitoring parameters, match the growth parameters, physiological parameters and environmental parameters in the monitoring parameters, and match the matching parameters. After the monitoring parameters are normalized.

Based on the content of the above embodiment, as an optional embodiment, the context perception module 403 is used to fuse the monitoring parameters based on the deep learning model to obtain regional context information; based on the regional context information, obtain the overall perception result of the greenhouse; The overall perception result of the greenhouse is matched with the preset regulation decision to obtain decision information.

Based on the content of the foregoing embodiment, as an optional embodiment, the decision-making information at least includes environmental regulation information, water and fertilizer management information, and agricultural operation information.

Based on the content of the foregoing embodiment, as an optional embodiment, the deep learning model is obtained by training based on a wide and deep neural network.

In the device provided by the embodiment of the present invention, the monitoring parameters of greenhouse crops are obtained, and the monitoring parameters are preprocessed. Based on the deep learning model, the preprocessed monitoring parameters are subjected to situational awareness to obtain decision information, and production management is carried out based on the decision information. By sensing multiple growth, physiological indicators and environmental parameters of crops and making corresponding decisions, the traditional greenhouse crop management method can be changed, so as to realize the intelligent production, monitoring and regulation of greenhouse crops, thereby improving the utilization rate of greenhouse monitoring data.

Secondly, by applying context-aware methods and concepts to greenhouse crop production, context-aware decision-making can be realized from the perspective of greenhouse crops, and the production management of greenhouse crop feedback can be realized at a lower cost. In addition, due to the modular design of the situation awareness framework, it can be combined according to different greenhouse environments and conditions, which is convenient for future maintenance and development, and then realizes multi-level integration to meet more complex production needs. Finally, because the deep learning model is used as the inference engine for situational perception, according to the structure and characteristics of the deep network, more types of situational information such as numerical values, texts, images, etc. can be processed at the same time, and the perception results are accurate and fast, which can meet the needs of the system. required timeliness and robustness.

FIG. 5 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG. 5 , the electronic device may include: a processor (processor) 510, a communication interface (Communications Interface) 520, a memory (memory) 530 and a communication bus 540, The processor 510 , the communication interface 520 , and the memory 530 communicate with each other through the communication bus 540 . The processor 510 can call the logic instructions in the memory 530 to execute the following method: obtain monitoring parameters of the greenhouse crops, and preprocess the monitoring parameters, the monitoring parameters include at least growth parameters, physiological parameters and environmental parameters; based on the deep learning model, Perform situational awareness on the preprocessed monitoring parameters to obtain decision-making information, and conduct production management based on the decision-making information. It should be noted that, in actual implementation, the electronic device may be in the form of a PC or a tablet computer. The PC or the tablet computer can collect data and have a decision-making control function, which is not specifically limited in this embodiment of the present invention.

In addition, the above-mentioned logic instructions in the memory 530 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, an electronic device, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

Embodiments of the present invention further provide a non-transitory computer-readable storage medium on which a computer program is stored, and the computer program is implemented by a processor to execute the methods provided in the above-mentioned embodiments, for example, including: obtaining the data of greenhouse crops. Monitoring parameters, and preprocessing the monitoring parameters. The monitoring parameters include at least growth parameters, physiological parameters and environmental parameters; based on the deep learning model, the preprocessed monitoring parameters are contextually perceived to obtain decision-making information, and production is based on the decision-making information. manage.

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic A disc, an optical disc, etc., includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (7)

1. A greenhouse crop context-aware production management method is characterized by comprising the following steps:

acquiring monitoring parameters of greenhouse crops, and preprocessing the monitoring parameters, wherein the monitoring parameters at least comprise growth parameters, physiological parameters and environmental parameters;

based on a deep learning model, performing context awareness on the preprocessed monitoring parameters to obtain decision information, and performing production management based on the decision information;

the method for obtaining the decision information by performing context awareness on the preprocessed monitoring parameters based on the deep learning model comprises the following steps:

based on a deep learning model, fusing monitoring parameters to obtain regional scene information;

providing the overall perception result of the greenhouse by using the scene information of each region according to a majority principle in a minority;

matching the greenhouse integral sensing result with a preset regulation and control decision to obtain decision information;

before the preprocessing of the monitoring parameters, the method further comprises the following steps:

screening the monitoring parameters based on the type of the greenhouse crop;

the decision information at least comprises environment regulation and control information, water and fertilizer management information and farming operation information.

2. The method according to claim 1, wherein the growth parameters comprise at least crop growth parameters, plant height and canopy leaf area parameters; the physiological parameters at least comprise crop net photosynthetic rate, canopy temperature, transpiration and crop fluorescence parameters; the environmental parameters at least comprise air temperature, air humidity, soil temperature, soil humidity, illumination intensity and CO₂And (4) concentration.

3. The method of claim 1, wherein the pre-processing the monitored parameter comprises:

and removing the missing value parameters in the monitoring parameters, matching the growth parameters, the physiological parameters and the environmental parameters in the monitoring parameters, and normalizing the matched monitoring parameters.

4. The method of any one of claims 1 to 3, wherein the deep learning model is based on wide-deep neural network training.

5. A greenhouse crop context aware production management device, comprising:

the system comprises an acquisition module, a monitoring module and a control module, wherein the acquisition module is used for acquiring monitoring parameters of greenhouse crops, and the monitoring parameters at least comprise growth parameters, physiological parameters and environmental parameters;

the preprocessing module is used for preprocessing the monitoring parameters;

the context awareness module is used for performing context awareness on the preprocessed monitoring parameters based on a deep learning model to obtain decision information;

the production management module is used for carrying out production management based on the decision information;

a screening module for screening the monitoring parameters based on the type of the greenhouse crop;

the context awareness module is further used for fusing the monitoring parameters based on the deep learning model to obtain regional context information; providing the overall perception result of the greenhouse by using the scene information of each region according to a majority principle in a minority; and matching the greenhouse integral sensing result with a preset regulation and control decision to obtain the decision information.

6. An electronic device, comprising:

at least one processor; and

at least one memory communicatively coupled to the processor, wherein:

the memory stores program instructions executable by the processor, the processor invoking the program instructions to perform the method of any of claims 1 to 4.

7. A non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the method of any one of claims 1 to 4.

Images