Method for constructing database of agronomic parameters and agricultural machinery parameters of pepper planting

Abstract

The invention belongs to the technical field of agricultural informatization, and provides a pepper planting agronomic parameter and agricultural machinery parameter database construction method, which comprises the following steps: deploying soil moisture content, illumination intensity and plant growth sensors in a pepper planting field, and collecting environmental parameters and agronomic requirements in different growth periods in real time; and key parameters during operation of the agricultural machine are associated. Edge computing nodes are used for carrying out space-time dimension dynamic mapping on agronomic data and model parameters, an environment-agronomic-model ternary association database is formed, and the problem that static parameters of a traditional database cannot adapt to a complex field environment is solved.

Description

Technical Field

[0001] This invention belongs to the field of agricultural information technology, specifically relating to a method for constructing a database of agronomic parameters and agricultural machinery parameters for chili pepper cultivation. Background Technology

[0002] Currently, agronomic schemes and machinery selection for processing chili pepper cultivation largely rely on experience-based matching, lacking systematic data support. Existing databases mostly statically store machinery parameters or agronomic indicators, failing to reflect the dynamic adaptation relationship between the two under different environmental conditions. This leads to low machinery operation efficiency and large fluctuations in chili pepper quality (e.g., fruit breakage exceeding 15% due to improper harvester parameters). Therefore, there is an urgent need for a database construction method that can dynamically link agronomic and machinery parameters. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a method for constructing a database of agronomic and agricultural machinery parameters for chili pepper cultivation, thereby resolving the issues in the prior art. The technical solution adopted by this invention is as follows: A method for constructing a database of agronomic and agricultural machinery parameters for chili pepper cultivation includes the following steps: Step 1: Modular acquisition of agronomic parameters; Step 2: Standardized input of agricultural machinery parameters; Step 3, Dynamic Association Model Construction; Step 4: Database self-update.

[0004] Furthermore, step 1 includes: dividing the processing chili peppers into four modules according to their growth period: seedling raising, transplanting, plant protection, and harvesting; each module collects core agronomic parameters through a sensor array, including: for example, the seedling raising module collects germination temperature and light duration; the harvesting module collects fruit maturity and pedicel toughness; the transplanting module collects transplanting time and transplanting density; and the plant protection module collects pest and disease types and pesticide usage.

[0005] Furthermore, step 2 includes: for the agricultural machinery corresponding to each module, inputting the agricultural machinery parameters and marking the effective threshold of the agricultural machinery parameters.

[0006] Furthermore, step 3 includes: constructing an association model based on a Bayesian network, with agronomic parameters as input nodes and environmental factors as latent variables, and agricultural machinery parameters as output nodes and environmental factors as latent variables, to construct a dynamic association model; training the dynamic association model through the EM algorithm to generate conditional probability tables for agronomic parameters, environmental factors, and agricultural machinery parameters.

[0007] Furthermore, the dynamic association model can be expressed as the following formula: ; in, , which are agronomic parameters; , which are environmental factors; These are agricultural machinery parameters; , are the conditional probability table parameters of a Bayesian network; Let i be the i-th agronomic parameter; Here are the probability distribution parameters corresponding to environmental factor E; Given an environmental factor E, the i-th agronomic parameter The conditional probability distribution parameters; Given agronomic parameter A and environmental factor E, the j-th agricultural machinery parameter The conditional probability distribution parameters; It is an index of agricultural machinery parameters; This is the total number of agricultural machinery parameters; Indicates in the parameter The joint probability distribution of agronomic parameter A, environmental factor E, and agricultural machinery parameter M is shown below.

[0008] Furthermore, in step 3, a dynamic correlation model is trained using the EM algorithm to generate conditional probability tables for agronomic parameters, environmental factors, and agricultural machinery parameters, including the following steps: Step 3.1, calculate the posterior probability of the latent environmental factors, using the following formula: ; in, It is the posterior probability distribution of the latent variables; It is the set of model parameters at the t-th iteration; Step 3.2: Based on the posterior probability results, maximize the expected log-likelihood function, as shown in the formula: ; in, It is the expected log-likelihood function; Step 3.3: Repeat steps 3.1-3.2 until convergence, then output the conditional probability table of A, E, and M.

[0009] Furthermore, step 4 includes: periodically collecting actual operation data, correcting model parameters through gradient descent, and realizing dynamic iteration of the database.

[0010] Furthermore, step 4 includes the following steps: Step 4.1: Regularly collect actual operational data on chili pepper cultivation. s is the data sample number; S is the total number of samples of actual operation data; Step 4.2, Define the loss function The formula is: ; in, It is the predicted j-th agricultural machinery parameter of the s-th sample. These are the actual agricultural machinery parameters used in operation; Step 4.3: Minimize the loss function using gradient descent. Update parameters The formula is: ; in, It's the learning rate. It is the loss function with respect to the current parameters The gradient; Step 4.4: Repeat steps 4.1-4.3 to achieve dynamic iteration of the database.

[0011] The present invention has the following beneficial effects: The database constructed by this invention can achieve precise matching between chili processing agronomy and machine parameters, effectively improving the efficiency of agricultural machinery operations and improving chili quality indicators (such as breakage rate and spiciness stability). At the same time, it can adapt to the environmental differences in different production areas and has strong practicality and scalability. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of the overall flow of the present invention. Detailed Implementation

[0013] The following will be based on embodiments of the present invention. Figure 1 The technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0014] like Figure 1 A method for constructing a database of agronomic and agricultural machinery parameters for chili pepper cultivation includes the following steps: Step 1: Modular acquisition of agronomic parameters; The process of processing chili peppers is divided into four modules: seedling raising, transplanting, plant protection, and harvesting. Each module collects core agronomic parameters through a sensor array (such as germination temperature (15-30℃) and light duration (8-12h / d) in the seedling raising module; and fruit maturity (coloring rate ≥90%) and pedicel toughness (≥5N) in the harvesting module).

[0015] This invention breaks down chili pepper cultivation into four core modules: seedling raising, transplanting, plant protection, and harvesting. Each module corresponds to specific agronomic indicators (such as germination rate in the seedling raising module and planting depth in the transplanting module). An independent parameter sub-library is developed for each module, recording the performance parameters (such as operating efficiency and damage rate) of different agricultural machinery models (such as semi-automatic and fully automatic transplanters) when matching specific agronomic indicators. Dynamic calls between sub-libraries are achieved through a module interface protocol, overcoming the shortcomings of traditional databases in handling changes in the combination of agronomic stages.

[0016] This invention utilizes sensors deployed in chili pepper fields to monitor soil moisture, light intensity, and plant growth, collecting real-time environmental parameters and agronomical requirements (such as irrigation, fertilization, and plant spacing) at different growth stages (seedling, transplanting, and harvesting). These parameters are then correlated with key parameters used in agricultural machinery operations (such as the row spacing adjustment range of the seeder and the cutting height of the harvester). Edge computing nodes are used to dynamically map agronomical data to machine parameters across time and space, forming a ternary relational database of "environment-agronomy-machinery," thus solving the problem that traditional databases with static parameters cannot adapt to complex field environments.

[0017] This invention rationally deploys soil moisture sensors, light intensity sensors, temperature and humidity sensors, and plant growth sensors in different areas of a chili pepper growing field (such as seedling area, transplanting area, and harvesting area) to ensure comprehensive and accurate collection of environmental and agronomical data. Data transmission and processing: Wireless communication technologies (such as LoRa and NB-IoT) are used to transmit the data collected by the sensors to edge computing nodes. The edge computing nodes preprocess the data, including data cleaning, format conversion, and outlier removal, and then perform spatiotemporal correlation analysis to establish a mapping relationship between environmental parameters, agronomic requirements, and agricultural machinery parameters. Database construction and updating: A dynamic database is built based on the correlation analysis results. The database has the function of receiving new data in real time and automatically updating the mapping relationship to ensure the timeliness and accuracy of the data.

[0018] This invention divides planting areas and installs various sensors in designated locations, debugging them to ensure proper functioning and accurate data acquisition. Edge computing nodes are configured with parameters, including data preprocessing rules and correlation analysis algorithm parameters. The sensors transmit collected data to the edge computing nodes in real time, where they preprocess the data according to preset rules. The edge computing nodes then use correlation analysis algorithms to analyze the preprocessed data, establishing a mapping relationship between "environment-agronomy-machine type" and storing it in a dynamic database. As the planting process progresses, the sensors continuously collect data and transmit it to the edge computing nodes, which continuously perform correlation analysis and update the mapping relationships in the database.

[0019] Based on the growth cycle and planting process of processed chili peppers, this invention scientifically and rationally divides the planting process into four modules: seedling raising, transplanting, plant protection, and harvesting, and clarifies the boundaries and core agronomic indicators of each module.

[0020] Step 2: Standardized entry of machine model parameters; For each module of agricultural machinery (such as seedling machine and transplanter), enter the key parameters (such as the seedling machine's tray size (50 holes / 72 holes) and the transplanter's plant spacing adjustment range (20-40cm)) and mark the effective threshold of the parameters (such as the optimal speed range of the harvester 200-300r / min).

[0021] Step 3: Construction of the dynamic association model; A Bayesian network-based association model is constructed, with agronomic parameters as input nodes, machine parameters as output nodes, and environmental factors (soil pH and precipitation) as latent variables. The model is trained using the EM algorithm to generate a conditional probability table of "agronomic parameters-environmental factors-machine parameters" (example: when soil pH = 6.5-7.0 and pepper stalk toughness = 5-8N, the fit probability of a harvester speed of 250 r / min reaches 92%).

[0022] This invention uses quality indicators (spiciness, moisture content) of processed chili varieties (such as long peppers and sweet peppers) as target variables, and planting agronomic parameters (such as planting density and harvesting cycle) as input features. It then uses a random forest model to inversely deduce optimal agricultural machinery parameters (such as harvester speed and conveyor belt inclination angle). The database has a built-in self-learning module that iteratively updates model weights based on measured data from different production areas, improving cross-regional adaptability.

[0023] Specifically, the dynamic association model is represented by the following formula: ; in, These are agronomic parameters, such as planting density, harvesting cycle, germination temperature, and transplanting density. These are environmental factors, such as soil moisture and diurnal temperature range; These are agricultural machinery parameters; , are the conditional probability table parameters of a Bayesian network; Let i be the i-th agronomic parameter; Here are the probability distribution parameters corresponding to environmental factor E; Given an environmental factor E, the i-th agronomic parameter The conditional probability distribution parameters; Given agronomic parameter A and environmental factor E, the j-th agricultural machinery parameter The conditional probability distribution parameters; It is an index of agricultural machinery parameters; This is the total number of agricultural machinery parameters; Indicates in the parameter The joint probability distribution of agronomic parameter A, environmental factor E, and agricultural machinery parameter M is shown below.

[0024] Furthermore, in step 3, a dynamic correlation model is trained using the EM algorithm to generate conditional probability tables for agronomic parameters, environmental factors, and agricultural machinery parameters, including the following steps: Step 3.1, calculate the posterior probability of the latent environmental factors, using the following formula: ; in, It is the posterior probability distribution of the latent variables; It is the set of model parameters at the t-th iteration; Step 3.2: Based on the posterior probability results, maximize the expected log-likelihood function, as shown in the formula: ; in, It is the expected log-likelihood function; Step 3.3: Repeat steps 3.1-3.2 until convergence, then output the conditional probability table of A, E, and M.

[0025] Step 4: Database self-update; Collect actual operation data every quarter (such as the damage rate of peppers after agricultural machinery operation), and correct model parameters through gradient descent to achieve dynamic iteration of the database.

[0026] Step 4 includes the following steps: Step 4.1: Regularly collect actual operational data on chili pepper cultivation. s is the data sample number; S is the total number of samples of actual operation data; Step 4.2, Define the loss function The formula is: ; in, It is the predicted j-th agricultural machinery parameter of the s-th sample. These are the actual agricultural machinery parameters used in operation; Step 4.3: Minimize the loss function using gradient descent. Update parameters The formula is: ; in, It's the learning rate. It is the loss function with respect to the current parameters The gradient; Step 4.4: Repeat steps 4.1-4.3 to achieve dynamic iteration of the database.

[0027] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Any modifications, alterations, alterations, or substitutions made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A database construction method of agronomic parameters and agricultural machine parameters for pepper cultivation, characterized by, Comprise the following steps: Step 1, agricultural parameter modular collection; Step 2, agricultural machinery parameter standardized input; Step 3, dynamic correlation model construction; Step 4, database self-update.

2. The method of claim 1, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Step 1 includes: according to the processing of pepper growth period is divided into seedling, transplanting, plant protection, harvesting four modules; each module through the sensor array to collect core agricultural parameters, including: such as seedling module to collect germination temperature, light duration; harvesting module to collect fruit maturity, fruit stem toughness; transplanting module to collect transplanting time, transplanting density; plant protection module to collect the types of pests and diseases, pesticide usage.

3. The method of claim 2, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Step 2 includes: for each module corresponding agricultural machinery, input agricultural machinery parameters, and mark the effective threshold of agricultural machinery parameters.

4. The method of claim 1, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Step 3 includes: based on Bayesian network correlation model is constructed, taking agricultural parameters as input nodes, environmental factors as hidden variables, agricultural machinery parameters as output nodes, environmental factors as hidden variables, to build a dynamic correlation model; through the EM algorithm to train dynamic correlation model, generate agricultural parameters, environmental factors, agricultural machinery parameters of conditional probability table.

5. The method of claim 4, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Dynamic correlation model is expressed as the following formula: ； in, , which are agronomic parameters; , which are environmental factors; These are agricultural machinery parameters; , are the conditional probability table parameters of a Bayesian network; Let i be the i-th agronomic parameter; Here are the probability distribution parameters corresponding to environmental factor E; Given an environmental factor E, the i-th agronomic parameter The conditional probability distribution parameters; Given agronomic parameter A and environmental factor E, the j-th agricultural machinery parameter The conditional probability distribution parameters; It is an index of agricultural machinery parameters; This is the total number of agricultural machinery parameters; Indicates in the parameter The joint probability distribution of agronomic parameter A, environmental factor E, and agricultural machinery parameter M is shown below.

6. The method of claim 5, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. In step 3, the EM algorithm is used to train the dynamic correlation model, and the conditional probability table of agricultural parameters, environmental factors and agricultural machinery parameters is generated, including the following steps: Step 3.1, calculate the posterior probability of hidden variable environmental factor, the formula is: ； wherein, is the posterior distribution of the latent variables; is the set of model parameters at the tth iteration; Step 3.2, based on the results of posterior probability, maximize the expected log-likelihood function, the formula is: ； wherein is the desired log-likelihood function; Step 3.3, repeat step 3.1-step 3.2 to convergence, output A, E, M conditional probability table.

7. The method of claim 6, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Step 4 includes: regularly collect actual operation data, correct the model parameters by gradient descent method, realize the dynamic iteration of database.

8. The method of claim 7, wherein the database is constructed by using the agricultural parameters and the agricultural machine parameters of the pepper cultivation. Step 4 includes the following steps: Step 4.1, collecting actual operation data of pepper planting regularly s is the data sample serial number; S is the total number of samples of actual operation data; Step 4.2, define loss function , the formula is: ； wherein, is the jth agricultural machine parameter of the sth sample predicted, is the agricultural machine parameter of the actual operation; Step 4.3, minimize the loss function by gradient descent , update parameters , formula: ； wherein, is the learning rate, is the gradient of the loss function with respect to the current parameters ; Step 4.4, repeat step 4.1-step 4.3, realize the dynamic iteration of database.

Images