Parameter determination method, system, device and storage medium for a feeder device

Abstract

The embodiment of the application provides a parameter determination method, system and device of a feeding device and a storage medium, and belongs to the technical field of agricultural machinery design. The method comprises the following steps: acquiring input information, wherein the input information comprises a particle type of a target particle processed by the feeding device, an optimization target and a constraint condition; acquiring particle parameters corresponding to the particle type according to the particle type; generating at least one group of initial structure parameters corresponding to the feeding device according to the particle parameters and the constraint condition; performing simulation experiments according to the at least one group of initial structure parameters to obtain experimental data; analyzing the experimental data to obtain an initial performance index corresponding to each group of initial structure parameters; and iteratively optimizing the at least one group of initial structure parameters and the at least one group of initial performance indexes according to the optimization target and the constraint condition to obtain at least one group of target structure parameters, wherein the parameters of the feeding device comprise the at least one group of target structure parameters. The embodiment of the application can realize intelligent design and optimization of the feeding device.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural machinery design, and in particular to a parameter determination method, system and device of a feeding device and a storage medium. BACKGROUND

[0002] The roller mechanical feeding device is a core component of a seeding machine and a sorting device, and is used for uniform and orderly feeding. The performance of the roller mechanical feeding device directly affects the seeding quality (such as uniformity, qualified rate, missing grain rate and heavy grain rate) and the sorting operation efficiency. At present, the design process of the roller mechanical feeding device highly depends on the experience of engineers, and usually needs to go through multiple iterations of “design-manufacture-test-modification”. The cycle is long, the cost is high, and it is difficult to ensure that the optimal structure parameters and working parameters are achieved.

[0003] At present, although computer-aided design (CAD) and discrete element simulation (DEM) technologies have been applied to the auxiliary design of the roller mechanical feeding device, the structure design and the simulation analysis process are independent of each other, there is a lack of effective integration between the links, and the steps of parameter setting, simulation execution, result statistics and analysis, and optimization decision usually need to be completed manually. Moreover, it is difficult to automatically adjust and optimize according to different particle characteristics, and the degree of automation is low. SUMMARY

[0004] The main purpose of the embodiments of the present application is to provide a parameter determination method, system, device and storage medium of a feeding device, which can solve the problem that the design and optimization of the roller mechanical feeding device in the prior art rely on manual work and have a low degree of automation, and realize intelligent design of the roller mechanical feeding device from structure design, simulation analysis to parameter optimization.

[0005] To achieve the above purpose, a first aspect of the embodiments of the present application provides a parameter determination method of a feeding device, which comprises:

[0006] obtaining input information, wherein the input information comprises a particle type of a target particle processed by the feeding device, an optimization target and a constraint condition;

[0007] obtaining particle parameters corresponding to the particle type according to the particle type;

[0008] generating at least one group of initial structure parameters corresponding to the feeding device according to the particle parameters and the constraint condition;

[0009] performing a simulation experiment according to at least one group of the initial structure parameters to obtain experimental data;

[0010] analyzing the experimental data to obtain an initial performance index corresponding to each group of the initial structure parameters;

[0011] According to the optimization target and the constraint condition, at least one set of initial structure parameters and at least one set of initial performance indicators are iteratively optimized to obtain at least one set of target structure parameters, and the parameters of the feeding device include at least one set of the target structure parameters.

[0012] In some embodiments, the obtaining of the particle parameter corresponding to the particle type according to the particle type comprises:

[0013] According to the particle type, the particle parameter corresponding to the particle type is queried;

[0014] In the case that the particle parameter corresponding to the particle type cannot be queried, a prompt information is sent to the client, and the prompt information is used to prompt the user to input the particle parameter corresponding to the particle type;

[0015] In the case that the particle parameter corresponding to the particle type input by the user cannot be obtained, the particle parameter corresponding to a preset general particle type is used as the particle parameter corresponding to the particle type.

[0016] In some embodiments, the simulation experiment according to at least one set of initial structure parameters comprises:

[0017] For each set of initial structure parameters, a three-dimensional modeling software encapsulation layer is called to map the initial structure parameters to a three-dimensional modeling software script, and a three-dimensional model of the feeding device is generated or modified according to the three-dimensional modeling software script;

[0018] A discrete element simulation software encapsulation layer is called to generate a discrete element simulation software script according to the particle parameter, the initial structure parameter, the three-dimensional model and a preset simulation control parameter, and the simulation control parameter includes a simulation duration and a data output frequency;

[0019] According to the discrete element simulation software script, a simulation environment is configured and a simulation experiment is performed to obtain the experimental data.

[0020] In some embodiments, the initial performance indicators include an initial seeding qualified rate, an initial grain leakage rate and an initial grain weight rate, and the optimization target is to maximize the seeding qualified rate, minimize the grain leakage rate and minimize the grain weight rate.

[0021] According to the optimization target and the constraint condition, at least one set of initial structure parameters and at least one set of initial performance indicators are iteratively optimized to obtain at least one set of target structure parameters, and the parameters of the feeding device include at least one set of the target structure parameters.

[0022] In each iteration process, a multi-objective optimization algorithm is executed on at least one set of initial structure parameters and at least one set of initial performance indicators based on the optimization target to obtain a set of candidate structure parameters;

[0023] The candidate structure parameter is taken as the initial structure parameter, and the step of performing simulation experiment according to at least one set of the initial structure parameters to obtain experimental data is executed until an iteration stop condition is met, and at least one set of target structure parameters is obtained.

[0024] In some embodiments, the multi-objective optimization algorithm is performed on at least one set of the initial structure parameters and at least one set of the initial performance indicators based on the optimization target to obtain a set of candidate structure parameters, including:

[0025] At least one set of the initial structure parameters and at least one set of the initial performance indicators are taken as an initial population;

[0026] Selection, crossover or mutation operations are performed on the initial population to generate a set of temporary structure parameters;

[0027] If the temporary structure parameter meets the constraint condition, the temporary structure parameter is taken as the candidate structure parameter;

[0028] If the temporary structure parameter does not meet the constraint condition, the step of performing selection, crossover or mutation operations on the initial population to generate a set of temporary structure parameters is executed until the temporary structure parameter meets the constraint condition.

[0029] In some embodiments, the candidate structure parameter is taken as the initial structure parameter, and the step of performing simulation experiment according to at least one set of the initial structure parameters to obtain experimental data is executed until an iteration stop condition is met, and at least one set of target structure parameters is obtained, including:

[0030] In each iteration process, it is judged whether the performance indicator corresponding to the candidate structure parameter in the current iteration number meets an iteration stop condition, and the iteration stop condition is that the average change rate of the performance indicator corresponding to the candidate structure parameter in a continuous preset number of iterations is less than a preset threshold value, or the current iteration number reaches a preset maximum iteration number;

[0031] If the performance indicator corresponding to the candidate structure parameter in the current iteration number does not meet the iteration stop condition, the candidate structure parameter is taken as the initial structure parameter, and the step of performing simulation experiment according to at least one set of the initial structure parameters to obtain experimental data is executed;

[0032] If the performance indicator corresponding to the candidate structure parameter in the current iteration number meets the iteration stop condition, the iteration is stopped, and at least one set of the target structure parameters is selected from a plurality of sets of the candidate structure parameters obtained in the iteration process.

[0033] In some embodiments, after iteratively optimizing at least one set of the initial structure parameters and at least one set of the initial performance indicators according to the optimization target and the constraint condition, the method further comprises:

[0034] generating a design report of the feeding device according to the parameters of the feeding device, and sending the design report of the feeding device to the client.

[0035] To achieve the above object, a second aspect of the embodiments of the present application provides a parameter determination system of a feeding device, which comprises a user interface intelligent agent, a task management intelligent agent, a knowledge base management intelligent agent, a parameter generation intelligent agent, a three-dimensional modeling intelligent agent, a discrete element simulation intelligent agent, a result analysis intelligent agent and an optimization intelligent agent.

[0036] The user interface intelligent agent acquires input information, which comprises a particle type of a target particle processed by the feeding device, an optimization target and a constraint condition.

[0037] The task management intelligent agent acquires particle parameters corresponding to the particle type from the knowledge base management intelligent agent according to the particle type.

[0038] The parameter generation intelligent agent generates at least one set of initial structure parameters corresponding to the feeding device according to the particle parameters and the constraint condition.

[0039] The three-dimensional modeling intelligent agent and the discrete element simulation intelligent agent perform simulation experiments according to at least one set of the initial structure parameters to obtain experimental data.

[0040] The result analysis intelligent agent analyzes the experimental data to obtain initial performance indicators corresponding to each set of the initial structure parameters.

[0041] The optimization intelligent agent iteratively optimizes at least one set of the initial structure parameters and at least one set of the initial performance indicators according to the optimization target and the constraint condition to obtain at least one set of target structure parameters, and the parameters of the feeding device comprise at least one set of the target structure parameters.

[0042] To achieve the above object, a third aspect of the embodiments of the present application provides an electronic device, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the method of the first aspect when executing the computer program.

[0043] To achieve the above object, a fourth aspect of the embodiments of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method of the first aspect.

[0044] The parameter determination method, system, device and storage medium of the feeding device provided by the application, the user inputs the target particle type, the optimization target and the constraint condition through the user interface agent; the task management agent requests the particle parameters of the target particle type from the knowledge base management agent; the task management agent transmits the particle parameters and the constraint condition to the parameter generation agent to generate the initial mechanical feeding device structure parameters of the roller; the task management agent transmits the current mechanical feeding device structure parameters of the roller to the CAD modeling agent to automatically generate or modify the three-dimensional model; the task management agent transmits the three-dimensional model generated by the CAD modeling agent, the particle parameters and the structure parameters to the DEM simulation agent to automatically configure and execute the numerical simulation experiment of the seed arrangement process; the task management agent transmits the DEM simulation output data to the result analysis agent to calculate the initial performance indicators such as the seeding qualification rate, the particle missing rate and the particle weight rate; the task management agent transmits the performance indicators calculated by the result analysis agent and the initial structure parameters to the optimization agent to execute the multi-objective optimization algorithm for iteration and adjustment of the initial structure parameters; the result of each iteration is judged to see whether the optimization target is met or the maximum iteration number is reached; in the case where the iteration stopping condition is met, the optimal solution set of the multi-objective optimization algorithm is output as the target structure parameters to obtain the optimized mechanical feeding device structure parameters of the roller. The application realizes the intelligent closed-loop design of the feeding device parameters by automatically obtaining the particle parameters, generating the initial structure parameters, simulating the experiment and multi-objective iterative optimization, solves the problems of low efficiency and high cost caused by the traditional reliance on manual trial and error, and has the advantages of significantly improving the design accuracy and efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 is a flowchart of the parameter determination method of the feeding device provided by the embodiment of the application;

[0046] Figure 2 is a whole architecture block diagram of the parameter determination method of the feeding device provided by the embodiment of the application;

[0047] Figure 3 is a flowchart of the optimization decision work provided by the embodiment of the application;

[0048] Figure 4 is a hardware structure schematic diagram of the electronic device provided by the embodiment of the application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical scheme and advantages of the application more clear, the application is further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application and not to limit the application.

[0050] It should be noted that although the functional modules are divided in the device schematic diagram, and the logical sequence is shown in the flowchart, in some cases, the steps shown or described can be performed in a manner different from the module division in the device or the sequence in the flowchart. The terms "first", "second", etc. in the specification and claims and the above-described drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this application is for the purpose of describing embodiments of this application only, and is not intended to be limiting of this application.

[0052] The drum mechanical feeding device is widely used in seeding machines, sorting equipment and other agricultural and industrial equipment. As a core feeding component, it is mainly used to realize uniform and orderly feeding of materials. The structural design and working performance of the device directly affect the quality indicators of seeding operation (such as seeding uniformity, qualified rate, missing rate, and over-seeding rate) and the overall efficiency of sorting operation.

[0053] Currently, in the design process of the drum mechanical feeding device, it mainly relies on the experience judgment and manual trial-and-error method of engineering and technical personnel, and usually needs to go through multiple cycles of iteration of "structural design - prototype production - test verification - parameter adjustment", resulting in long research and development period, high cost, and difficulty in ensuring that the final design matches the optimal structural parameters and working performance.

[0054] Although computer-aided design (CAD) technology and discrete element simulation (DEM) methods have been applied in the design optimization of such devices, the existing technology still has the following shortcomings: CAD modeling, DEM simulation, data analysis and optimization are often separated, data exchange and process connection are not smooth, and there is a lack of effective integration between them; usually, parameter setting, simulation execution, result statistics and analysis, and optimization decision need to be completed manually, the process is tedious, and the iteration period is long; design adjustment of the drum mechanical feeding device for different crop particle characteristics still requires a lot of manual intervention and redesign; the knowledge and data accumulated in the design process are difficult to effectively deposit, reuse and intelligently apply; it is difficult to systematically explore a wide design space, and it is difficult to achieve multi-objective collaborative optimization based on experience.

[0055] Based on this, the embodiments of the present application provide a parameter determination method, system, device and storage medium for a feeding device, aiming to develop an intelligent design method capable of realizing the automation of the whole process from structural design, simulation analysis to parameter optimization, to improve the design efficiency and quality, adapt to the feeding needs of diversified material characteristics, and realize the intelligentization, systematization and integration of the design of the drum mechanical feeding device.

[0056] The parameter determination system of the feeding device provided by the embodiments of the present application mainly comprises a user interaction layer, a task coordination and management layer, an agent cooperation layer, a knowledge and data layer, and a tool integration layer. The core is that the agent cooperation layer comprises multiple functional agents, which collaboratively complete the automatic design and optimization of the drum mechanical feeding device through interaction with open source CAD software, open source DEM software, and an internal knowledge base.

[0057] Among them, the core agent and its functions are as follows:

[0058] User interface agent (UIA, User Interface Agent): used for receiving user input of particle characteristics, optimization objectives and constraint conditions, and displaying final design results.

[0059] Task management agent (TMA, Task Management Agent): as the system general dispatcher, used for task decomposition, agent scheduling, process control and progress management.

[0060] Knowledge base management agent (KBA, Knowledge Base Agent): used for maintaining a particle characteristics library and a drum mechanical feeding device design knowledge base, and providing decision support for other agents.

[0061] Parameter generation agent (PGA, Parameter Generation Agent): used for generating initial or iterative structure parameters and working parameters of the drum mechanical feeding device according to particle characteristics and the knowledge base.

[0062] CAD modeling agent (CMA, CAD Modeling Agent): used for interacting with open source CAD software (such as FreeCAD), automatically generating or modifying a three-dimensional model of the drum mechanical feeding device according to parameters provided by the PGA or optimization agent, and exporting the model into a format required by DEM simulation (such as STL).

[0063] DEM simulation agent (DSA, DEM Simulation Agent): used for interacting with open source DEM software (such as LIGGGHTS or Yade), automatically configuring and executing a numerical simulation experiment of the seed distribution process according to the model and related parameters provided by the CMA, and collecting simulation data.

[0064] Result analysis agent (RAA, Result Analysis Agent): used for analyzing simulation data output by the DSA, and calculating key performance indicators such as seeding qualification rate, grain loss rate, and grain weight rate.

[0065] Optimization Agent (OA): used to adjust the design parameters based on the analysis results of RAA, using multi-objective optimization algorithms (such as Non-dominated Sorting Genetic Algorithm II, Multi-objective Particle Swarm Optimization Algorithm, etc.), to guide the new round of design-modeling-simulation iteration until the Pareto optimal solution set is found or the user-defined convergence condition is met.

[0066] Report Generation Agent (RGA): used to integrate the information of the entire design optimization process and generate a technical scheme document.

[0067] The parameter determination method, system, equipment and storage medium of the feeding device provided by the embodiments of the application are specifically described through the following embodiments. First, the parameter determination method of the feeding device in the embodiments of the application is described.

[0068] The parameter determination method of the feeding device provided by the embodiments of the application relates to the technical field of agricultural machinery design. The parameter determination method of the feeding device provided by the embodiments of the application can be applied to a terminal, can be applied to a server end, and can also be software running in a terminal or a server end. In some embodiments, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc.; the server end can be configured as a stand-alone physical server, can be configured as a server cluster or a distributed system composed of multiple physical servers, can also be configured as a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and basic cloud computing services such as big data and artificial intelligence platforms; and the software can be an application that implements the parameter determination method of the feeding device, but is not limited to the above forms.

[0069] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0070] It should be noted that in various specific embodiments of the present application, when relevant processing needs to be performed on data related to the identity or characteristics of the user, such as user information, user behavior data, user history data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of such data will comply with relevant laws, regulations, and standards. In addition, when the embodiments of the present application need to obtain sensitive personal information of the user, the separate permission or separate consent of the user will be obtained through a pop-up window or a jump to a confirmation page, and after obtaining the separate permission or separate consent of the user, the necessary user-related data for the normal operation of the embodiments of the present application will be obtained.

[0071] Figure 1 is an optional flowchart of a parameter determination method of a feeding device provided by the embodiments of the present application, Figure 1 The method in the above step S100-S600 can include but is not limited to the steps S100-S600.

[0072] In step S100, input information is obtained, the input information including a particle type of a target particle processed by the feeding device, an optimization target, and a constraint condition.

[0073] In the present embodiment, the user can input the particle type of the target particle, the optimization target, and the constraint condition through a user interface. The parameter determination system of the feeding device obtains the input information of the user through a user interface agent (UIA). The particle type of the target particle is specified by the user for the particle type that needs to be processed by the drum mechanical feeding device, such as corn, soybeans, wheat, etc.; the optimization target is defined by the user for the performance index of the drum mechanical feeding device that the user hopes to optimize, such as seeding qualification rate, grain loss rate, heavy grain rate, etc., and the target value is set; the constraint condition is set by the user for the limitation condition that needs to be met in the design process of the drum mechanical feeding device, such as the maximum size of the drum, cost limitation, etc. These information will be passed to the task management agent (TMA), and the subsequent agent cooperation process will be triggered by the TMA, and finally a high-performance drum mechanical feeding device design scheme meeting the user's demand will be generated.

[0074] In addition, after receiving the user input information, the UIA also needs to be responsible for the structured processing of the user input. The UIA will uniformly convert the user's free text input, selection box selection, numerical input, etc. into a data format that can be clearly recognized and processed by the system internally. After completing the structured processing, the UIA will send the structured user request information to the TMA through an internal communication mechanism (such as message-based asynchronous communication). After receiving the structured user request from the UIA, the TMA creates a new design optimization task, assigns a unique task ID, and ensures task data isolation and traceability.

[0075] Exemplarily, in the optimization design of an air-roller for corn kernels, the user specifies the crop as "corn" through the UIA and inputs the typical characteristic parameters of corn kernels (or calls from the library by the KBA, such as the average particle size, three-axis size distribution, density, friction coefficient, moisture content, etc.). The user sets the optimization goal as maximizing the seeding qualified rate, minimizing the missing kernel rate, and minimizing the heavy kernel rate. The user sets the constraint condition as the roller diameter range value (such as 150-300 mm).

[0076] In step S200, the particle parameters corresponding to the particle type are obtained according to the particle type.

[0077] In this embodiment, when the user inputs the target particle type through the UIA, the UIA transmits this information to the TMA. As the overall scheduler of the system, the TMA immediately initiates a request to the knowledge base management agent (KBA) to obtain the detailed parameters of the type of particles after receiving the particle type information. The KBA can automatically retrieve or calculate the corresponding particle physical and mechanical parameters according to the particle type specified by the user, providing data support for subsequent optimization. These parameters can include but are not limited to the average particle size, particle size distribution (such as three-axis size distribution), particle shape description (such as sphericity, aspect ratio), particle density, particle moisture content in the physical parameters, and the friction coefficient between particles and the inner wall / type hole of the roller, the elastic modulus of the particle, the Poisson's ratio, etc. in the mechanical parameters. These parameter data can come from public literature, standard database, historical project accumulation, or be manually input by the user in specific cases. In addition, the KBA also includes a roller mechanical feeder design rule library, which stores information about the design principles, empirical rules, theoretical formulas, and literature conclusions of the roller mechanical feeder. These information are associated with particle parameters and device performance indicators.

[0078] Specifically, the knowledge base includes a particle characteristic database and a roller mechanical feeder design rule library. The particle characteristic database is used to store the physical (size, shape, density, moisture content, etc.) and mechanical (friction coefficient, elastic modulus, angle of repose, etc.) parameters of different particles. The roller mechanical feeder design rule library is used to store the empirical rules, theoretical formulas, and literature conclusions of the relationship between the roller mechanical feeder type (mechanical, pneumatic), key structural parameters (type hole size, shape, number, air pressure, etc.), and performance indicators. The KBA is responsible for managing this knowledge base, and the PGA queries this knowledge base when generating initial parameters, and the OA can also refer to the constraint or heuristic information in it during the optimization process.

[0079] In this embodiment, upon receiving the request from TMA, KBA queries its maintained database of particle properties. If the database contains a complete record of the parameters of the particle type, KBA extracts these standardized parameter data and returns them to TMA. If KBA fails to find a complete set of parameters of the particle type specified by the user in the database, KBA can send a prompt to the user through TMA and UIA, asking the user to manually input the required particle parameters. If the user is unable to provide all the parameters, KBA can return a default or general parameter template for such a particle type.

[0080] For example, the user inputs the target particle type as “corn” through UIA, and UIA sends the “corn” type information to TMA. Then TMA requests the parameters of “corn” particles from KBA, and KBA queries its database of particle properties and finds and returns the typical parameters of corn, such as: average particle size 8 mm, size distribution range 6-10 mm, density 1.25 g / cm 3 and so on. After receiving these parameters, TMA passes them to the parameter generation agent (PGA) for subsequent generation of initial drum parameters.

[0081] Step S300: generating at least one set of initial structural parameters corresponding to the feeding device according to the particle parameters and the constraint conditions.

[0082] In this embodiment, after receiving the particle parameters and the constraint conditions, the parameter generation agent (PGA) queries the drum mechanical feeding device design rule base maintained by KBA, which stores a large amount of design experience, theoretical formulas, literature conclusions, and association rules between particle properties and device parameters. For example, the rule base can include guidelines such as “the hole diameter should be slightly larger than 1.2-1.5 times the average particle size”, “the drum rotation speed should be inversely proportional to the particle size and proportional to the density”, “the number of holes should be related to the drum diameter and the expected seed rate”, etc. PGA applies these rules to provide guidance for the generation of initial parameters based on the current particle parameters and constraint conditions.

[0083] Specifically, based on the understanding of the particle parameters, the constraint conditions, and the application of the design rules, PGA uses a pre-set parameter generation logic or algorithm to calculate and generate at least one set of initial structural parameters. This process can be rule-based reasoning, simple mathematical calculation, or preliminary heuristic optimization. The generated parameters can include but are not limited to: drum diameter, thickness, hole size / number / layout, air chamber parameters.

[0084] For example, PGA receives the parameters of corn particles (such as average particle size 8 mm, size distribution 6-10 mm, density 1.25 g / cm 3After the design rules of KBA are queried from the design rule base, the rules about the size of the corn granular type hole, the relationship between the roller speed and the particle size are found out. PGA performs calculation and reasoning according to the rules and constraints. For example, according to the particle size distribution, the range of the type hole diameter is preliminarily determined to be 10-15mm; according to the density and particle size, the roller diameter is preliminarily recommended to be an intermediate value in the constraint range, such as 200mm; according to the roller diameter and the expected seed rate, the number and layout mode of the type holes (such as 2 rows, 5 in each row, and the center distance of 40mm) are preliminarily determined. PGA generates one or more sets of initial parameter combinations. PGA returns the generated initial parameter set to TMA, and TMA transmits it to the CAD modeling agent (CMA) to start the preliminary CAD modeling process.

[0085] In step S400, simulation experiments are performed according to at least one set of initial structure parameters to obtain experimental data.

[0086] In this embodiment, TMA transmits at least one set of initial structure parameters generated by PGA to the three-dimensional modeling agent (i.e., the CAD modeling agent, CMA) and the discrete element simulation agent (i.e., the DEM simulation agent, DSA). CMA automatically generates or modifies a three-dimensional geometric model of the roller mechanical feed device according to these parameters. After receiving the target particle parameters, the initial structure parameters, and the geometric model file provided by CMA, DSA calls the DEM software to perform simulation calculation. The DEM software simulates the entire process of roller rotation, particle movement, type hole filling and release on the computer according to the input parameters. After the simulation is completed, DSA extracts the required data from the output file of the DEM software to obtain the experimental data.

[0087] For example, TMA transmits the initial parameter set generated by PGA (such as a roller diameter of 200mm, a type hole diameter of 12mm, a 2-row 5-column layout, a speed of 100rpm, and an air suction pressure of 3000Pa) to CMA and DSA. CMA generates an STL model of the roller and transmits it to DSA. After receiving the STL model, the corn particle parameters (such as an average particle size of 8mm, a size distribution of 6-10mm, a density of 1.25g / cm 3 For example, TMA transmits the initial parameter set generated by PGA (such as a roller diameter of 200mm, a type hole diameter of 12mm, a 2-row 5-column layout, a speed of 100rpm, and an air suction pressure of 3000Pa) to CMA and DSA. CMA generates an STL model of the roller and transmits it to DSA. After receiving the STL model, the corn particle parameters (such as an average particle size of 8mm, a size distribution of 6-10mm, a density of 1.25g / cm

[0088] In step S500, the experimental data is analyzed to obtain an initial performance index corresponding to each set of initial structure parameters.

[0089] In this embodiment, the result analysis agent (RAA) receives the simulation data from the DSA, and calculates the initial performance indicators such as the seeding qualification rate, the missing seed rate, the double seed rate, the seed damage evaluation (based on the maximum contact force or stress between seeds, if the model supports), the seeding rate, and the uniformity coefficient of the plant spacing, etc. Then, the RAA organizes the calculated performance indicators into structured data. These data, together with the corresponding initial structure parameters, are returned to the TMA.

[0090] For example, the TMA transmits the simulation data (including the trajectories of the corn seeds, the filling state of the type hole, the seed drop points, etc.) completed by the DSA to the RAA. The RAA calculates the number of corn seeds falling into the preset seeding area according to the preset analysis logic, and calculates the qualification rate. The RAA analyzes whether the type hole is empty when it reaches the seeding area, and calculates the missing seed rate. The RAA analyzes the time interval of the seed drop points, and judges and calculates the double seed rate. According to the calculation results, the qualification rate is 85%, the missing seed rate is 5%, and the double seed rate is 3%.

[0091] In step S600, at least one group of initial structure parameters and at least one group of initial performance indicators are iteratively optimized according to the optimization target and the constraint condition, to obtain at least one group of target structure parameters. The parameters of the feeding device include at least one group of target structure parameters.

[0092] In this embodiment, the TMA transmits the optimization target and the constraint condition set by the user, as well as the initial structure parameters and the corresponding initial performance indicators, to the optimization agent (OA). The OA uses a multi-objective optimization algorithm (such as the non-dominated sorting genetic algorithm II, the multi-objective particle swarm optimization algorithm, etc.) integrated in the system in advance to iteratively process the data. In each iteration, the algorithm generates a new group of candidate parameters. For these new groups of parameters, the OA instructs the TMA to start a new round of design-modeling-simulation process. After each iteration, the OA evaluates whether the optimization process converges. When the convergence condition is met, the optimization process terminates. After the optimization terminates, the OA outputs one or more groups of parameters representing the Pareto optimal solution set as the target structure parameters. These groups of parameters achieve a good balance between multiple user-defined objectives while satisfying all constraints. For example, one group of parameters may result in a high qualification rate but a slightly high missing seed rate, and another group of parameters may result in a low missing seed rate but a slightly low qualification rate. The user can select the most suitable solution according to actual needs. In this way, the system can automatically start from the initial design, iteratively optimize, and ultimately obtain high-quality target structure parameters that meet the user's multi-objective requirements and satisfy the constraint conditions.

[0093] Exemplarily, the TMA transmits initial structure parameters (such as a roller diameter of 200 mm, a hole diameter of 12 mm, and a layout of 2 rows and 5 columns) and initial performance indicators (a qualified rate of 85%, a missing particle rate of 5%, and a heavy particle rate of 3%) to the OA. The OA uses a non-dominated sorting genetic algorithm II (NSGA-II) algorithm to generate a new parameter group by performing selection, crossover, and mutation operations on the initial structure parameters as an initial population according to the optimization target (maximizing the qualified rate, minimizing the missing particle rate, and minimizing the heavy particle rate) and the constraint condition (the roller diameter range is 100-300 mm) set by the user. For each newly generated parameter group, the OA instructs the TMA to start a new round of modeling, simulation, and analysis processes to obtain new performance indicators. The OA reevaluates the population according to the new structure parameters and performance indicators, updates, and continues iteration. After multiple iterations, the performance of the individuals in the population tends to be stable, the Pareto frontier no longer changes significantly, and the convergence condition is reached. The OA outputs a Pareto optimal solution set, for example: scheme A: a roller diameter of 220 mm, a hole diameter of 11 mm, a qualified rate of 92%, a missing particle rate of 2.5%, and a heavy particle rate of 2.8%; and scheme B: a roller diameter of 190 mm, a hole diameter of 13 mm, a qualified rate of 90%, a missing particle rate of 1.8%, and a heavy particle rate of 3.5%. The TMA takes one or more of the schemes (for example, the user may prefer the scheme B with a lower missing particle rate) as the target structure parameters.

[0094] The embodiment eliminates the manual operation link in the traditional design method by integrating three-dimensional modeling, discrete element simulation, and multi-objective optimization algorithm, greatly improves the parameter optimization efficiency. The system can automatically adapt to different particle types, quickly generate and evaluate multiple sets of structure parameters, and effectively explore a wider design space. The closed-loop optimization mechanism ensures that each parameter adjustment is based on accurate simulation results, improving the reliability and performance of the final design scheme. In addition, by constructing and updating the knowledge base, the system can continuously accumulate design experience, provide intelligent decision support for future similar projects, further shorten the design cycle, and improve the design quality.

[0095] In some embodiments, step S200 can include but is not limited to steps S210 to S230:

[0096] Step S210, querying the particle parameters corresponding to the particle type according to the particle type;

[0097] Step S220, in the case that the particle parameters corresponding to the particle type cannot be queried, sending a prompt information to the client, the prompt information being used to prompt the user to input the particle parameters corresponding to the particle type;

[0098] Step S230: If the particle parameters corresponding to the particle type input by the user cannot be obtained, the particle parameters corresponding to the preset general particle type are used as the particle parameters corresponding to the particle type.

[0099] In this embodiment, KBA first attempts to query the maintained particle characteristic database based on the particle type specified by the user. If KBA cannot find a parameter record in the database that perfectly matches the user-specified particle type, KBA will not directly cause the system to fail or return an error. Instead, KBA will send a prompt message to the user interface. The prompt message may include input boxes for key parameters such as particle density, particle size distribution, and friction coefficient. After KBA sends the prompt message, the system will attempt to obtain the parameters entered by the user. If the user does not enter any parameters within a preset time or the entered parameters are incomplete, KBA will automatically select the particle parameters corresponding to a preset general particle type as the default value. The preset general particle type may refer to a representative type of particle defined in the knowledge base (e.g., "general grains," "general seeds," etc.), and its parameters are set to typical or intermediate values ​​applicable to that type of particle. KBA will assign these general parameters to the particle type currently specified by the user but lacking in the knowledge base, thereby enabling the subsequent design process to continue.

[0100] This embodiment employs a multi-layered parameter acquisition strategy. When parameters cannot be directly obtained from the system, accurate parameters are acquired through user interaction, enhancing the system's adaptability. By introducing preset universal parameters as a fallback solution, it ensures that the parameter acquisition process can output valid data regardless of whether the particle type is known, thus achieving automation and intelligence in the particle parameter acquisition process.

[0101] In some embodiments, step S400 may include, but is not limited to, steps S410 to S430:

[0102] Step S410: For each set of initial structural parameters, call the 3D modeling software encapsulation layer to map the initial structural parameters to a 3D modeling software script, and generate or modify the 3D model of the feeding device according to the 3D modeling software script.

[0103] Step S420: Call the encapsulation layer of the discrete element simulation software to generate a discrete element simulation software script based on the particle parameters, the initial structural parameters, the three-dimensional model and the preset simulation control parameters. The simulation control parameters include simulation duration and data output frequency.

[0104] Step S430: Configure the simulation environment and conduct simulation experiments according to the discrete element simulation software script to obtain the experimental data.

[0105] In this embodiment, TMA passes at least a set of initial structure parameters (e.g. drum diameter, orifice size, layout, etc.) generated by PGA to CMA. After receiving these parameters, CMA first calls the Wrapper of the 3D modeling software. This Wrapper is a pre-developed or integrated interface component of the system, which wraps the underlying API of open-source 3D modeling software (e.g. FreeCAD, OpenCASCADE, etc.) so that the agent can interact with these software in a unified way without concerning the specific implementation details inside the software. CMA utilizes this Wrapper to map the received initial structure parameters into script instructions or data structures that the Wrapper can understand and pass to the underlying CAD software. Subsequently, CMA drives the underlying CAD software to automatically generate or modify the 3D model of the drum mechanical feeder based on these scripts or data, and exports it into a format suitable for subsequent DEM simulation (e.g. STL, STEP, etc.).

[0106] In this embodiment, after the 3D model generation or modification is completed, CMA passes the model file, particle parameters, initial structure parameters, and pre-set simulation control parameters (e.g. simulation duration, data output frequency, gravitational acceleration, solver parameters, etc.) to DSA. After receiving these information, DSA calls the Wrapper of the discrete element simulation software. This Wrapper wraps the input file generation logic or API of open-source discrete element simulation software (e.g. Yade, LIGGGHTS, etc.). DSA utilizes this Wrapper to automatically generate the input script file required by the simulation software based on the received particle parameters, structure parameters, 3D model, and simulation control parameters. The script will contain: importing the STL model to define the drum geometry, generating particles in the specified area according to the parameters, setting the contact model between particles and particles, particles and the drum, defining the rotational motion of the drum, setting the gravity, setting the data output frequency, etc.

[0107] In this embodiment, after DSA generates the simulation script, it calls the Wrapper to execute the script. This is equivalent to starting the simulation calculation of the underlying DEM software. According to the script configuration, the simulation software simulates the entire process of drum rotation, particle filling, orifice filling and release, particle discharge, etc. in the virtual environment. During the simulation process, according to the output frequency set in the script, the DEM software records the raw data such as the position, velocity, force of each particle at each time step, the orifice filling state, etc. and saves it to the output file. DSA monitors the simulation process to ensure its normal completion and collects these output files. After the simulation is completed, DSA extracts the data from the output file of the DEM software to obtain the experimental data.

[0108] The embodiment integrates CAD modeling and DEM simulation under a unified system framework, and data transmission and process connection are smooth; from parameters to model, and then to simulation configuration and execution, the whole process is automatically completed by an intelligent agent without human intervention, which greatly improves the efficiency; through encapsulation layer technology, the universality and scalability of the system are enhanced, so that different modeling software and simulation software can be flexibly replaced and integrated; in addition, the preset simulation control parameters ensure the consistency and comparability of the simulation experiment, and provide a reliable data basis for subsequent result analysis and optimization.

[0109] In some embodiments, step S600 can include but is not limited to steps S610 to S620:

[0110] Step S610, in the process of each iteration, a multi-objective optimization algorithm is performed on at least one set of initial structure parameters and at least one set of initial performance indicators based on the optimization target, to obtain a set of candidate structure parameters, the initial performance indicators include initial sowing qualified rate, initial grain loss rate and initial grain weight, and the optimization target is to maximize the sowing qualified rate, minimize the grain loss rate and minimize the grain weight.

[0111] Step S620, the candidate structure parameters are taken as the initial structure parameters, and the step of performing simulation experiment according to at least one set of initial structure parameters to obtain experimental data is executed until the iteration stopping condition is met, to obtain at least one set of target structure parameters.

[0112] In the embodiment, the TMA transmits the initial structure parameters generated by the PGA and the initial performance indicators (sowing qualified rate, grain loss rate, grain weight, etc.) calculated by the RAA to the OA. The OA iteratively optimizes the initial structure parameters and the initial performance indicators according to the optimization target (such as maximizing the qualified rate, minimizing the grain loss rate, and minimizing the grain weight) and the constraint conditions set by the user.

[0113] Specifically, in each iteration process, OA executes a multi-objective optimization algorithm according to the current initial structure parameters and initial performance indicator data, in combination with the optimization objectives and constraint conditions set by the user. Commonly used algorithms include non-dominated sorting genetic algorithm II (NSGA-II), multi-objective particle swarm optimization algorithm, etc. These algorithms can handle the case where multiple objectives exist and may conflict, aiming to find one or more Pareto optimal solutions, i.e., the best state set reached when it is impossible to improve one objective without sacrificing at least one other objective. The multi-objective optimization algorithm generates a new set of candidate structure parameters based on the current structure parameters and performance indicator data. For example, the algorithm may attempt to slightly increase the size of a certain type of hole to try to reduce the grain loss rate while monitoring the impact on the qualified rate and heavy grain rate; or try to adjust the layout of the type of hole to balance multiple indicators. This newly generated set of candidate structure parameters is passed to TMA as the "initial structure parameters" in the next iteration. TMA regards it as a new design scheme and starts a new round of modeling-simulation-analysis process (i.e., TMA passes the new "initial structure parameters" to CMA for modeling, then to DSA for simulation, and to RAA for analysis to obtain new "initial performance indicators". The new "parameter + indicator" set is passed to OA again for the next iteration step). If the current iteration meets the iteration stopping condition, the optimization process ends, and OA outputs a set of Pareto optimal solutions as the optimization result, i.e., one or more sets of target structure parameters. These parameter sets have achieved a good balance between multiple objectives under all constraint conditions.

[0114] Through the multi-objective optimization algorithm, the embodiment can simultaneously consider multiple performance indicators such as seeding qualified rate, grain loss rate, and heavy grain rate, and find the optimal combination of structure parameters under the premise of meeting the constraint conditions; the iterative optimization process has high automation degree, reduces manual intervention, and improves parameter optimization efficiency; simulation experiments are used instead of physical experiments, which shortens the development cycle and reduces costs; by setting reasonable iteration stopping conditions, the convergence and reliability of the optimization result are guaranteed, so that the final target structure parameters can better meet the performance requirements of the feeding device.

[0115] In some embodiments, step S610 can include but is not limited to steps S611 to S614:

[0116] Step S611, taking at least one set of the initial structure parameters and at least one set of the initial performance indicators as an initial population;

[0117] Step S612, performing selection, crossover, or mutation operations on the initial population to generate a set of temporary structure parameters;

[0118] Step S613, if the temporary structure parameter satisfies the constraint condition, the temporary structure parameter is taken as the candidate structure parameter;

[0119] Step S614, if the temporary structure parameter does not satisfy the constraint condition, jump to the step of performing selection, crossover or mutation operation on the initial population to generate a set of temporary structure parameters until the temporary structure parameter satisfies the constraint condition.

[0120] In this embodiment, at least one set of initial structure parameters generated by PGA according to particle characteristics and knowledge base, and at least one set of initial performance indicators calculated by RAA according to simulation data output by DSA, are taken as the initial population of the multi-objective optimization algorithm. Iterative calculation of the multi-objective optimization algorithm is performed on the initial population.

[0121] Specifically, in each iteration of the algorithm, selection, crossover and / or mutation operations are performed on individuals in the current population according to the rules of the multi-objective optimization algorithm. These operations aim to simulate the natural evolution process, explore new design space, and produce offspring individuals with potentially better performance. The selection operation tends to retain individuals with better performance; the crossover operation combines the parameters of two or more individuals; and the mutation operation randomly perturbs the parameters of an individual. Through these operations, a set of new temporary structure parameters is generated, which represents the design candidates of the next generation. It is judged whether the newly generated temporary structure parameters satisfy the constraint conditions. These constraint conditions can include but are not limited to: maximum / minimum diameter of the drum, maximum / minimum size of the hole, pressure range of the air chamber, upper limit of manufacturing cost, etc. physical or design restrictions. If the temporary structure parameter satisfies the constraint condition, the set of temporary structure parameters is confirmed as a set of valid candidate structure parameters. This set of candidate parameters will be used for the next round of design iteration, i.e. passed to CMA for new three-dimensional model generation. If the temporary structure parameter does not satisfy the constraint condition, the set of temporary structure parameters is discarded, and the step of generating a set of temporary structure parameters by performing selection, crossover or mutation operation on the initial population is jumped back to, to generate new temporary structure parameters again, and the constraint check is performed again until the temporary structure parameter that satisfies the constraint condition is generated.

[0122] This embodiment can effectively handle multiple conflicting objectives and fully explore the design space through the multi-objective optimization algorithm, increasing the possibility of obtaining the global optimal solution. By enforcing constraint checking during optimization, it ensures that the generated candidate structure parameters not only optimize performance, but also meet the requirements of engineering practice, avoiding ineffective or infeasible design exploration. It can flexibly adapt to different constraint condition settings and is suitable for various design optimization scenarios of mechanical drum feeders.

[0123] In some embodiments, step S620 can include, but is not limited to, steps S621-S623:

[0124] In step S621, in each iteration process, it is determined whether the performance index corresponding to the candidate structure parameter in the current iteration number meets the iteration stop condition, which is that the average change rate of the performance index corresponding to the candidate structure parameter in a continuous preset number of iterations is less than a preset threshold, or the current iteration number reaches a preset maximum iteration number.

[0125] In step S622, if the performance index corresponding to the candidate structure parameter in the current iteration number does not meet the iteration stop condition, the candidate structure parameter is taken as the initial structure parameter, and the step of performing simulation experiment according to at least one set of initial structure parameters to obtain experimental data is executed.

[0126] In step S623, if the performance index corresponding to the candidate structure parameter in the current iteration number meets the iteration stop condition, the iteration is stopped, and at least one set of target structure parameters is selected from the multiple sets of candidate structure parameters obtained in the iteration process.

[0127] In this embodiment, in each iteration process, it is determined whether the performance index corresponding to the candidate structure parameter in the current iteration number meets the iteration stop condition. If the iteration stop condition is not met, it indicates that the optimization process still has potential, and the current candidate structure parameter (as the initial structure parameter of this round) still has room for improvement or needs to explore new design directions. At this time, the candidate structure parameter obtained in this round of iteration is directly taken as the initial structure parameter of the next round of iteration, and then the step of performing simulation experiment according to at least one set of initial structure parameters to obtain experimental data is jumped back to start a new round of iteration cycle. If the iteration stop condition is met, it indicates that the optimization process has converged, or the preset iteration upper limit has been reached. At this time, the further iteration cycle is stopped. Then, at least one set of target structure parameters is selected from all the multiple sets of candidate structure parameters generated in the entire iteration optimization process. The selection criteria can be various, for example, selecting a single scheme with the best performance, selecting a set of non-inferior solution schemes constituting the Pareto front for user selection, or selecting according to other specific rules (such as the lowest comprehensive cost or the smallest manufacturing difficulty). This screening work can be automatically completed by the optimization agent (OA) according to the preset rules, or a recommended list is generated for the user to make the final decision. The selected target structure parameter is the final design optimization result.

[0128] Specifically, the iteration stopping condition includes that in a preset number of consecutive times (for example, 5 or 10 consecutive iterations), the average change rate of each key performance indicator corresponding to the candidate structure parameter is less than a preset threshold, that is, the Pareto frontier improvement rate is less than a preset threshold, which can be set to 0.1% or 1%, indicating that the performance improvement or deterioration has been very small, and the optimization process tends to converge. And the current cumulative number of iterations has reached the maximum number of iterations set by the user in advance, limiting the number of iterations to prevent the optimization process from circulating indefinitely.

[0129] The embodiment sets reasonable iteration stopping conditions, avoids ineffective excessive iteration, and improves optimization efficiency; selects target structure parameters from multiple candidate structure parameters, increasing the likelihood of obtaining high-quality solutions; automatically uses candidate structure parameters generated during the optimization process as the starting point for the next iteration, and intelligently decides whether to continue iteration or terminate optimization through continuous simulation experiments and performance evaluation according to the preset iteration stopping condition, thereby efficiently approaching or reaching the optimal design scheme.

[0130] In some embodiments, step S600 can include but is not limited to step S700:

[0131] Step S700, generating a design report of the feeding device according to the parameters of the feeding device, and sending the design report of the feeding device to the client.

[0132] In the embodiment, when the OA completes the iterative optimization and outputs the parameters of the feeding device to the TMA, the TMA summarizes the key data related to the final target structure parameters generated during the entire design optimization process. These data can include target structure parameters (such as optimized roller diameter, thickness, hole size / number / layout, air chamber parameters, etc.), performance indicators (such as seeding qualification rate, grain loss rate, heavy grain rate, plant spacing uniformity coefficient of variation, etc.), constraint conditions, optimization process summary (such as the type of optimization algorithm used, total number of iterations, convergence, etc.), feeding device type (such as air suction type, mechanical type, etc.), particle characteristics, key model / simulation file information (such as CAD model file information corresponding to the final optimization scheme or link or embedded content of key simulation result charts), etc.

[0133] In this embodiment, TMA transmits the aggregated data to the Report Generation Agent (RGA). RGA automatically generates a detailed design report based on a preset report template (which may include standardized chapters, formats, chart styles, etc.) and the received data. The design report may include the following: task overview (feeding device type, design objectives, constraints, etc.), particle characteristics (including key physical and mechanical parameters of the particles used), design parameters (optimized drum structure parameters and operating parameters), performance prediction (key performance indicators under the optimized scheme and their comparison with target values), a brief description of the optimization process (briefly explaining the optimization algorithm, iteration process, and results), design drawings or model links (screenshots of the CAD model of the final optimized scheme, key views, or links to model files), simulation result charts (such as particle motion trajectory diagrams, cavity filling diagrams, performance indicator curves changing with iteration counts, etc.), conclusions and recommendations (summarizing design results and possibly including suggestions for subsequent manufacturing, testing, or improvement), etc.

[0134] In this embodiment, after RGA generates a report, the report and related files can be sent to the customer via UIA, or the report content can be displayed directly on the UIA interface.

[0135] This embodiment eliminates the tedious process of manually writing design reports by automating report generation, avoiding human error and ensuring the accuracy and consistency of report content. The automatic report generation and sending function enables designers to quickly obtain optimization results and process information, facilitating result analysis and decision-making.

[0136] like Figure 2 As shown, the complete implementation of this embodiment is as follows, wherein the optimization decision-making process is as follows: Figure 3 As shown:

[0137] S1: Users input the type / name of the target particles through UIA, as well as the desired optimization goals (such as seeding qualification rate, particle leakage rate, and target value of heavy particle rate) and constraints (such as the maximum size of the drum mechanical feeding device, cost limits, etc.).

[0138] S2: UIA structures the user request and sends it to TMA.

[0139] S3: TMA creates new design optimization tasks, assigns a unique task ID, and ensures task data isolation and traceability.

[0140] S4: TMA requests detailed physical and mechanical property parameters for the specified particle type from KBA.

[0141] S5: KBA queries knowledge base, returns normalized granule property data (e.g. size distribution, density, friction coefficient, elastic modulus, etc.). If no such granule information is found in the knowledge base, KBA can prompt TMA to ask user for input of granule property data (e.g. size distribution, density, friction coefficient, elastic modulus, etc.) or start from a pre-set general template (default template for crop category in KBA is invoked when user cannot provide all property data).

[0142] S6: TMA passes granule properties and user constraints (e.g. preference for drum mechanical feeder type) to PGA.

[0143] S7: PGA generates one or more sets of initial drum mechanical feeder key structural parameters (e.g. drum diameter, thickness, orifice size / number / layout, plenum parameters, etc.) by combining KBA-provided drum mechanical feeder design principles and empirical rules (e.g. recommended orifice range for specific granule size, air pressure range, etc.).

[0144] S8: PGA returns generated parameter scheme to TMA.

[0145] S9: TMA obtains current drum mechanical feeder structural parameter set from PGA (first iteration) or OA (subsequent optimization iterations).

[0146] S10: TMA issues modeling task to CMA, passing structural parameters and output requirements (e.g. STEP for archiving in traditional modeling, STL for better handling in DEM numerical simulation).

[0147] S11: CMA invokes 3D modeling software wrapper, maps design parameters to 3D modeling software script, and automatically generates or modifies 3D model based on parameters.

[0148] S12: CMA returns generated model file path (or file itself, depending on implementation) to TMA.

[0149] S13: TMA issues simulation task to DSA, passing: STL model file generated by CMA, detailed granule DEM parameters provided by KBA (based on granule properties), current drum mechanical feeder operating parameters provided by PGA / OA (e.g. drum rotational speed, air flow pressure, etc.), simulation control parameters (e.g. simulation duration, data output frequency).

[0150] S14: DSA invokes DEM software (Yade / LIGGGHTS) wrapper, automatically generates DEM input script, configures simulation environment (imports geometry, generates granules, sets contact model, defines boundaries and motions, sets solver), and executes DEM simulation.

[0151] S15: DSA monitors the simulation process, collects raw output data (particle trajectories, forces, velocities, contacts, etc.) and log files.

[0152] S16: DSA returns the simulation completion status and output file path / initial summary information to TMA.

[0153] S17: TMA issues a result analysis task to RAA, passing the data path of DSA output.

[0154] S18: RAA loads and processes DEM simulation data, calculates key performance indicators such as: seeding eligibility rate (based on particle landing position and time interval in the target area), grain leakage rate, heavy grain rate, particle damage assessment (based on maximum contact force or stress between particles if the model supports), seeding rate, plant spacing uniformity coefficient of variation, etc.

[0155] S19: RAA returns the quantified performance indicators and analysis summary to TMA.

[0156] S20: TMA passes the performance indicators obtained by RAA analysis and the corresponding design parameters to OA.

[0157] S21: OA executes multi-objective optimization algorithms (such as NSGA-II, MOPSO, etc.) according to the preset optimization objectives (such as maximizing eligibility rate, minimizing grain leakage / heavy grain rate) and constraint conditions.

[0158] S22: OA generates a new set of candidate design parameters within the parameter optimization range using a diversity preservation operator.

[0159] S23: OA sends the new parameter set to TMA, and TMA triggers a new CAD modeling→DEM simulation→result analysis iteration.

[0160] S24: If the Pareto front improvement rate is less than the threshold value (user-defined, the target function change rate of the optimal solution after n consecutive iterations) or the number of iterations reaches the user-defined maximum number of iterations, i.e., if the optimization converges or reaches the maximum number of iterations, OA outputs the Pareto optimal solution set or recommends the best design parameter scheme to TMA.

[0161] S25: TMA passes the final optimization design parameters, related CAD models, key DEM simulation results, optimization process data, etc. to RGA.

[0162] S26: RGA writes detailed technical scheme documents according to the preset templates and user requirements, describing the optimized drum mechanical feed device design, performance prediction, etc.

[0163] S27: RGA returns the generated documents to TMA.

[0164] S28: The TMA presents the final technical scheme document and other related results (such as key model files, simulation animation links, etc.) to the user through the UIA.

[0165] S29: If the user does not make further modification requests, the task is archived and ends; otherwise, the relevant parameters are re-iterated according to the user's modification requests.

[0166] S30: The task is completed, and the TMA archives all task data and logs.

[0167] The embodiments of the present application define the clear division of labor of each agent, the asynchronous communication mechanism and the collaboration protocol based on messages through the multi-agent collaboration mechanism, and are centrally coordinated by the TMA to ensure process automation and efficient collaboration; through the deep integration with open source CAD / DEM software, the automatic call, parameterized control and result extraction of the agent to FreeCAD and LIGGGHTS / Yade and the like are realized through a special interface encapsulation layer; the initial parameter generation based on knowledge uses the particle characteristic data stored in the KBA and the design principle of the drum mechanical feeding device to intelligently generate a reasonable initial design parameter range for different particles, thereby improving the optimization efficiency; the automated multi-objective optimization integrates multi-objective optimization algorithms, can simultaneously optimize multiple conflicting seed arrangement performance indicators, and obtains a set of Pareto optimal solutions for the user to select; the closed-loop feedback and adaptive design form a closed-loop feedback of "parameter generation-modeling-simulation-analysis-optimization", can automatically adjust the design parameters according to the simulation results, and realize the adaptive design for different particle characteristics. The embodiments of the present application greatly shorten the design cycle of the drum mechanical feeding device through multi-agent collaboration and automated processes, and reduce manual intervention; can systematically perform multi-objective optimization, explore a wider design space, and obtain a drum mechanical feeding device design scheme with better performance, rather than just local improvement; the system can automatically adjust and optimize the structure parameters of the drum mechanical feeding device according to the input specific particle characteristics (physical and mechanical parameters), adapt to the seeding needs of different crops; the constructed knowledge base can continuously accumulate design experience and simulation data, and is applied to the subsequent design process through the agent to realize the sedimentation and intelligent use of knowledge; the CAD modeling, DEM simulation, data analysis, multi-objective optimization and the like are closely integrated under the multi-agent framework to realize an end-to-end automated design optimization process; the number of trial production of physical prototypes and the test cost are reduced, and the design optimization is driven through simulation; the automated system can explore non-traditional or counter-intuitive design parameter combinations, which may give rise to a drum mechanical feeding device with novel structure and better performance.

[0168] The embodiment of the present application further provides a parameter determination system of a feeding device, the parameter determination system of the feeding device comprising a user interface intelligent agent, a task management intelligent agent, a knowledge base management intelligent agent, a parameter generation intelligent agent, a three-dimensional modeling intelligent agent, a discrete element simulation intelligent agent, a result analysis intelligent agent and an optimization intelligent agent;

[0169] The user interface intelligent agent acquires input information, and the input information comprises a particle type of a target particle processed by the feeding device, an optimization target and a constraint condition;

[0170] The task management intelligent agent acquires, according to the particle type, particle parameters corresponding to the particle type from the knowledge base management intelligent agent;

[0171] The parameter generation intelligent agent generates at least one set of initial structure parameters corresponding to the feeding device according to the particle parameters and the constraint condition;

[0172] The three-dimensional modeling intelligent agent and the discrete element simulation intelligent agent perform simulation experiments according to at least one set of the initial structure parameters to obtain experimental data;

[0173] The result analysis intelligent agent analyzes the experimental data to obtain an initial performance index corresponding to each set of the initial structure parameters;

[0174] The optimization intelligent agent performs iterative optimization on at least one set of the initial structure parameters and at least one set of the initial performance index according to the optimization target and the constraint condition to obtain at least one set of target structure parameters, and the parameters of the feeding device comprise at least one set of the target structure parameters.

[0175] In some embodiments, the task management intelligent agent acquires, according to the particle type, particle parameters corresponding to the particle type from the knowledge base management intelligent agent, comprising:

[0176] The task management intelligent agent requests the knowledge base management intelligent agent for particle parameters corresponding to the particle type according to the particle type;

[0177] The knowledge base management intelligent agent queries the particle parameters corresponding to the particle type and sends the particle parameters to the task management intelligent agent;

[0178] In a case where the particle parameters corresponding to the particle type cannot be queried, the task management intelligent agent sends prompt information to a user through the user interface intelligent agent, and the prompt information is used to prompt the user to input the particle parameters corresponding to the particle type through the user interface intelligent agent;

[0179] In a case where the granule parameters corresponding to the particle type input by the user cannot be obtained, the knowledge base management agent takes the preset granule parameters corresponding to a general particle type as the granule parameters corresponding to the particle type.

[0180] In some embodiments, the three-dimensional modeling agent and the discrete element simulation agent perform simulation experiments according to at least one set of the initial structure parameters to obtain experimental data, including:

[0181] For each set of the initial structure parameters, the three-dimensional modeling agent calls a three-dimensional modeling software encapsulation layer, maps the initial structure parameters to a three-dimensional modeling software script, and generates or modifies a three-dimensional model of the feeding device according to the three-dimensional modeling software script.

[0182] The task management agent sends the granule parameters, the initial structure parameters, preset simulation control parameters, and the three-dimensional model to the discrete element simulation agent, and the simulation control parameters include a simulation duration and a data output frequency.

[0183] The discrete element simulation agent calls an encapsulation layer of a discrete element simulation software, and generates a discrete element simulation software script according to the granule parameters, the initial structure parameters, the simulation control parameters, and the three-dimensional model.

[0184] The discrete element simulation agent configures a simulation environment and performs a simulation experiment according to the discrete element simulation software script to obtain the experimental data.

[0185] In some embodiments, the optimization agent iteratively optimizes at least one set of the initial structure parameters and at least one set of the initial performance indicators according to the optimization target and the constraint condition to obtain at least one set of target structure parameters, including:

[0186] In each iteration, the optimization agent executes a multi-objective optimization algorithm based on the optimization target on at least one set of the initial structure parameters and at least one set of the initial performance indicators to obtain a set of candidate structure parameters, and the initial performance indicators include an initial seeding qualified rate, an initial grain loss rate, and an initial grain weight rate, and the optimization target is to maximize the seeding qualified rate, minimize the grain loss rate, and minimize the grain weight rate.

[0187] The optimization agent takes the candidate structure parameters as the initial structure parameters, jumps to the step of performing simulation experiments by the three-dimensional modeling agent and the discrete element simulation agent according to at least one set of the initial structure parameters to obtain experimental data, until an iteration stop condition is met, and at least one set of target structure parameters is obtained.

[0188] In some embodiments, the optimization agent performs a multi-objective optimization algorithm on at least one set of the initial structure parameters and at least one set of the initial performance indicators based on the optimization target to obtain a set of candidate structure parameters, including:

[0189] The optimization agent takes at least one set of the initial structure parameters and at least one set of the initial performance indicators as an initial population;

[0190] The optimization agent performs selection, crossover or mutation operations on the initial population to generate a set of temporary structure parameters;

[0191] If the temporary structure parameters meet the constraint condition, the optimization agent takes the temporary structure parameters as the candidate structure parameters;

[0192] If the temporary structure parameters do not meet the constraint condition, the optimization agent jumps to the step of performing selection, crossover or mutation operations on the initial population to generate a set of temporary structure parameters until the temporary structure parameters meet the constraint condition.

[0193] In some embodiments, the optimization agent takes the candidate structure parameters as the initial structure parameters, and jumps to the step of performing simulation experiments by the three-dimensional modeling agent and the discrete element simulation agent based on at least one set of the initial structure parameters to obtain experimental data until an iteration stopping condition is met to obtain at least one set of target structure parameters, including:

[0194] In each iteration process, the optimization agent judges whether the performance indicator corresponding to the candidate structure parameters in the current iteration number meets an iteration stopping condition, and the iteration stopping condition is that the average change rate of the performance indicator corresponding to the candidate structure parameters in a continuous preset number of iterations is less than a preset threshold, or the current iteration number reaches a preset maximum iteration number;

[0195] If the performance indicator corresponding to the candidate structure parameters in the current iteration number does not meet the iteration stopping condition, the optimization agent takes the candidate structure parameters as the initial structure parameters, and jumps to the step of performing simulation experiments by the three-dimensional modeling agent and the discrete element simulation agent based on at least one set of the initial structure parameters to obtain experimental data;

[0196] If the performance indicator corresponding to the candidate structure parameters in the current iteration number meets the iteration stopping condition, the iteration is stopped, and the optimization agent screens at least one set of the target structure parameters from a plurality of sets of the candidate structure parameters obtained from the iteration process.

[0197] In some embodiments, the parameter determination system of the feeding device further includes a report generation agent;

[0198] The report generation agent generates a design report of the feeding device according to the parameters of the feeding device;

[0199] The user interface agent sends the design report of the feeding device to a client.

[0200] The specific implementation of the parameter determination system of the feeding device is basically the same as the above-mentioned specific embodiments of the parameter determination method of the feeding device, and will not be repeated here.

[0201] The present application also provides an electronic device, which includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the above-mentioned parameter determination method of the feeding device. The electronic device can be any intelligent terminal including a tablet computer, a vehicle-mounted computer, etc.

[0202] Please refer to Figure 4 , Figure 4 The hardware structure of the electronic device of another embodiment is illustrated, which includes:

[0203] The processor 801 can be implemented in the form of a general central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits, etc., for executing related programs to realize the technical solutions provided by the present application;

[0204] The memory 802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM), etc. The memory 802 can store an operating system and other application programs, and when the technical solutions provided by the present application are implemented by software or firmware, the related program codes are saved in the memory 802 and called and executed by the processor 801 to realize the parameter determination method of the feeding device of the present application;

[0205] The input / output interface 803 is used to realize information input and output;

[0206] The communication interface 804 is used to realize the communication interaction between the device and other devices, which can realize communication through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.);

[0207] A bus 805 is used to transmit information between the various components (e.g., the processor 801, the memory 802, the input / output interface 803, and the communication interface 804) of the device.

[0208] The processor 801, the memory 802, the input / output interface 803, and the communication interface 804 are communicatively connected to each other within the device through the bus 805.

[0209] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program. The computer program is executed by a processor to implement the parameter determination method of the feeding device.

[0210] The memory, as a non-transitory computer readable storage medium, can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely arranged with respect to the processor, and these remote memories can be connected to the processor through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0211] The parameter determination method of the feeding device, the parameter determination device of the feeding device, the electronic equipment and the storage medium provided by the embodiments of the present application are as follows: a user inputs a target particle type, an optimization target and a constraint condition through a user interface agent; a task management agent requests particle parameters of the target particle type from a knowledge base management agent; the task management agent passes the particle parameters and the constraint condition to a parameter generation agent to generate initial mechanical feeding device structure parameters of a roller; the task management agent passes the current mechanical feeding device structure parameters of the roller to a CAD modeling agent to automatically generate or modify a three-dimensional model; the task management agent passes the three-dimensional model generated by the CAD modeling agent, the particle parameters and the structure parameters to a DEM simulation agent to automatically configure and execute a numerical simulation experiment of a seed spacing process; the task management agent passes DEM simulation output data to a result analysis agent to calculate performance indicators such as an initial seeding qualification rate, a particle missing rate and a particle overlapping rate; the task management agent passes the performance indicators calculated by the result analysis agent and the initial structure parameters to an optimization agent to execute a multi-objective optimization algorithm for iteration and adjustment of the initial structure parameters; it is determined whether the result of each iteration meets the optimization target or reaches a maximum iteration number; in the case where the iteration stopping condition is met, the optimal solution set of the multi-objective optimization algorithm is output as target structure parameters, and the optimized mechanical feeding device structure parameters of the roller are obtained. The present application realizes intelligent closed-loop design of the feeding device parameters by automatically obtaining particle parameters, generating initial structure parameters, simulating experiments and multi-objective iterative optimization, solves the problems of low efficiency and high cost caused by traditional reliance on manual trial and error, and has the advantages of significantly improving design accuracy and efficiency.

[0212] The embodiments described in the embodiments of the present application are used to more clearly illustrate the technical solutions of the embodiments of the present application, and do not constitute a limitation on the technical solutions provided by the embodiments of the present application. Those skilled in the art can know that, with the evolution of technology and the appearance of new application scenarios, the technical solutions provided by the embodiments of the present application are also applicable to similar technical problems.

[0213] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and can include more or fewer steps than those shown in the figures, or combine certain steps or different steps.

[0214] The device embodiments described above are only schematic, and units described as separate components can or can not be physically separate, that is, can be located in one place or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0215] Those skilled in the art can understand that all or some of the steps in the method disclosed above, the function modules / units in the system and the device can be implemented as software, firmware, hardware or appropriate combination thereof.

[0216] The terms "first", "second", "third", "fourth" etc. (if any) in the description and the drawings of the present application are used to distinguish similar objects, and do not necessarily indicate a particular order or sequence. It should be understood that the data thus used can be interchanged, where appropriate, so that the embodiments of the present application described herein can be carried out in other than the order shown or described herein. Furthermore, the terms "comprising" and "having", and any variations thereof, are intended to cover non-exclusive inclusion, for example, processes, methods, systems, products, or devices that comprise a list of steps or units not necessarily limited to those clearly listed, but can include other steps or units not clearly listed or inherent to such processes, methods, products, or devices.

[0217] It should be understood that in the present application, "at least one" means one or more, and "multiple" means two or more. "And / or" is used to describe the relationship between the associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that there are three cases: only A, only B, and A and B at the same time, where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can mean a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0218] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the above units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed objects can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0219] The units described as separate components above can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0220] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present alone, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0221] If the integrated unit is realized in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application, essentially or the part that contributes to the prior art, or all or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes multiple instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program storage media.

[0222] The preferred embodiments of the embodiments of the present application are described above with reference to the accompanying drawings, and the scope of the rights of the embodiments of the present application is not limited thereto. Any modifications, equivalent replacements and improvements made by those skilled in the art without departing from the scope and essence of the embodiments of the present application shall be within the scope of the rights of the embodiments of the present application.

Claims

1. A method for determining the parameters of a feeding device, characterized in that, The method includes: Obtain input information, which includes the particle type of the target particles processed by the feeding device, the optimization objective, and the constraints. Obtain the particle parameters corresponding to the particle type based on the particle type; Based on the particle parameters and the constraints, at least one set of initial structural parameters corresponding to the feeding device is generated; Simulation experiments were conducted based on at least one set of the initial structural parameters to obtain experimental data. The experimental data were analyzed to obtain the initial performance indicators corresponding to each set of initial structural parameters; Based on the optimization objective and the constraints, at least one set of initial structural parameters and at least one set of initial performance indicators are iteratively optimized to obtain at least one set of target structural parameters. The parameters of the feeding device include at least one set of target structural parameters. The initial performance indicators include the initial seeding qualification rate, the initial seed leakage rate, and the initial seed repetition rate. The optimization objective is to maximize the seeding qualification rate, minimize the seed leakage rate, and minimize the seed repetition rate. The step involves iteratively optimizing at least one set of initial structural parameters and at least one set of initial performance indices based on the optimization objective and the constraints to obtain at least one set of target structural parameters, including: During each iteration, a multi-objective optimization algorithm is executed on at least one set of initial structural parameters and at least one set of initial performance indicators based on the optimization objective to obtain a set of candidate structural parameters. The candidate structural parameters are used as the initial structural parameters, and the process jumps to the step of performing a simulation experiment based on at least one set of the initial structural parameters to obtain experimental data, until the iteration stopping condition is met, and at least one set of target structural parameters is obtained.

2. The method according to claim 1, characterized in that, The step of obtaining the particle parameters corresponding to the particle type based on the particle type includes: Query the particle parameters corresponding to the particle type based on the particle type; If the particle parameters corresponding to the particle type cannot be found, a prompt message is sent to the client, which prompts the user to input the particle parameters corresponding to the particle type. If the particle parameters corresponding to the particle type input by the user cannot be obtained, the particle parameters corresponding to the preset general particle type will be used as the particle parameters corresponding to the particle type.

3. The method according to claim 1, characterized in that, The simulation experiment based on at least one set of the initial structural parameters, to obtain experimental data, includes: For each set of initial structural parameters, the 3D modeling software encapsulation layer is invoked to map the initial structural parameters to a 3D modeling software script, and the 3D model of the feeding device is generated or modified according to the 3D modeling software script. The encapsulation layer of the discrete element simulation software is invoked to generate a discrete element simulation software script based on the particle parameters, the initial structural parameters, the three-dimensional model, and the preset simulation control parameters, including the simulation duration and data output frequency. Configure the simulation environment and conduct simulation experiments according to the discrete element simulation software script to obtain the experimental data.

4. The method according to claim 1, characterized in that, The multi-objective optimization algorithm is performed on at least one set of initial structural parameters and at least one set of initial performance indices based on the optimization objective to obtain a set of candidate structural parameters, including: Use at least one set of the initial structural parameters and at least one set of the initial performance indicators as the initial population; The initial population is subjected to selection, crossover, or mutation operations to generate a set of temporary structural parameters; If the temporary structural parameter satisfies the constraint condition, then the temporary structural parameter is used as the candidate structural parameter; If the temporary structure parameters do not meet the constraints, the process jumps to the step of performing selection, crossover, or mutation operations on the initial population to generate a set of temporary structure parameters, until the temporary structure parameters meet the constraints.

5. The method according to claim 1, characterized in that, The step of using the candidate structural parameters as the initial structural parameters and then proceeding to the step of performing simulation experiments based on at least one set of the initial structural parameters to obtain experimental data continues until the iteration stopping condition is met, resulting in at least one set of target structural parameters, including: In each iteration, it is determined whether the performance index corresponding to the candidate structure parameter in the current iteration number meets the iteration stopping condition. The iteration stopping condition is that the average rate of change of the performance index corresponding to the candidate structure parameter in a consecutive preset number of iterations is less than a preset threshold, or the current iteration number reaches the preset maximum number of iterations. If the performance index corresponding to the candidate structure parameter in the current iteration does not meet the iteration stopping condition, then the candidate structure parameter is used as the initial structure parameter, and the process jumps to the step of performing a simulation experiment based on at least one set of the initial structure parameters to obtain experimental data. If the performance index corresponding to the candidate structural parameters in the current iteration meets the iteration stopping condition, then the iteration stops, and at least one set of target structural parameters is selected from the multiple sets of candidate structural parameters obtained during the iteration process.

6. The method according to claim 1, characterized in that, After iteratively optimizing at least one set of initial structural parameters and at least one set of initial performance indices according to the optimization objective and the constraints to obtain at least one set of target structural parameters, the method further includes: A design report for the feeding device is generated based on the parameters of the feeding device, and the design report is sent to the client.

7. A parameter determination system for a feeding device, characterized in that, The parameter determination system of the feeding device includes a user interface agent, a task management agent, a knowledge base management agent, a parameter generation agent, a 3D modeling agent, a discrete element simulation agent, a result analysis agent, and an optimization agent; The user interface agent acquires input information, which includes the particle type, optimization objective, and constraints of the target particles processed by the feeding device. The task management agent obtains the particle parameters corresponding to the particle type from the knowledge base management agent according to the particle type; The parameter generating agent generates at least one set of initial structural parameters corresponding to the feeding device based on the particle parameters and the constraints. The 3D modeling agent and the discrete element simulation agent perform simulation experiments based on at least one set of initial structural parameters to obtain experimental data. The result analysis agent analyzes the experimental data to obtain the initial performance index corresponding to each set of initial structural parameters; The optimization agent iteratively optimizes at least one set of initial structural parameters and at least one set of initial performance indicators according to the optimization objective and the constraints, to obtain at least one set of target structural parameters. The parameters of the feeding device include at least one set of target structural parameters. The initial performance indicators include initial seeding qualification rate, initial seed leakage rate, and initial seed repetition rate. The optimization objective is to maximize the seeding qualification rate, minimize the seed leakage rate, and minimize the seed repetition rate. In each iteration, the optimization agent executes a multi-objective optimization algorithm on at least one set of initial structural parameters and at least one set of initial performance indicators based on the optimization objective, to obtain a set of candidate structural parameters. The candidate structural parameters are used as the initial structural parameters, and the process jumps to the step of performing a simulation experiment based on at least one set of initial structural parameters to obtain experimental data, until the iteration stopping condition is met, thus obtaining at least one set of target structural parameters.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the parameter determination method of the feeding device according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the parameter determination method of the feeding device according to any one of claims 1 to 6.

Images