A method for constructing a multi-moisture content corn kernel model based on discrete element method

Abstract

The present invention is applicable to the field of plant modeling and analysis technology and provides a method for constructing corn kernel models with multiple moisture contents based on the discrete element method. The method comprises the following steps: analyzing the geometric morphology and characteristic size parameters of corn kernels at maturity with a moisture content of 12% to 30% for the variety to be modeled, and determining the variation in the characteristic size of the corn kernels; obtaining the geometric parameters of a basic corn kernel model by analyzing the characteristic size parameters; predicting the volumetric size variation of kernels with different moisture contents using a sigmoid function in conjunction with a matrix equation; determining the arrangement of filling spheres for the basic corn kernel model based on the results of the geometric morphology analysis, calculating the coordinates of the filling spheres, and establishing a geometric model; determining the contact parameters of the basic model, and predicting the variation in surface adhesion behavior using a sigmoid function in conjunction with a matrix equation and a linear cohesion model to establish a humidity gradient coupled HM-LC model. The constructed model is consistent with the characteristics of actual corn kernel moisture content variation and is suitable for agricultural product processing and production processes.

Description

Technical Field

[0001] The present invention belongs to the technical field of plant modeling and analysis, and in particular relates to a method for constructing a multi-moisture content corn kernel model based on a discrete element method. Background Art

[0002] At the time of corn harvest, the moisture content of corn fluctuates significantly due to the influence of the surrounding environment. During the harvesting process, whether it is harvesting with a harvester or the subsequent threshing operation, the changes in the physical properties of the grains caused by the fluctuation of moisture content will make it difficult for the grains to adapt to mechanical action. Grains with too low a moisture content are easily broken by collision and extrusion, while grains with too high a moisture content are prone to sticking together during threshing, or uneven force will cause incomplete grain shedding, resulting in harvest losses. After entering the processing stage, corn with unstable moisture content will also increase the risk of breakage during transportation, storage and further processing due to differences in volume and surface state, affecting processing efficiency and product quality, and causing serious economic losses to corn harvesting and processing.

[0003] With the maturity of computer technology and finite element and discrete element theories, numerical simulation has become an important means to reveal the interaction mechanism between corn kernels. Compared with traditional field experiments that rely on the natural environment, this method has significant advantages such as efficient iteration, controllable costs, and breaking through seasonal restrictions. It is particularly suitable for analyzing the mechanical response of corn under complex working conditions such as mechanized harvesting, transportation, threshing, and cleaning. However, different numerical methods have obvious differences in modeling characteristics and applicable scenarios, and they need to be selected in a targeted manner based on the biomechanical properties of corn kernels. Therefore, in order to solve the above problems, there is an urgent need for a discrete element population model of corn kernels with different moisture contents that can be applied throughout the entire harvest process. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide a method for constructing a multi-moisture content corn kernel model based on the discrete element method, aiming to solve the problems raised in the above background technology.

[0005] The embodiment of the present invention is implemented as follows: a method for constructing a multi-moisture corn kernel model based on a discrete element method, comprising the following steps:

[0006] Step 1: Analyze the geometric morphology and characteristic size parameters of corn kernels at maturity with a moisture content of 12%-30% for the modeled variety, determine the changes in the characteristic size of corn kernels under different moisture contents, and select corn kernels with a moisture content of 12% as the basic model;

[0007] Step 2: According to the analysis results of the characteristic size parameters, the geometric parameters of the basic corn kernel model are obtained;

[0008] Step 3: Use the Sigmoid function with the matrix equation to predict the volume size changes of grains with different moisture contents;

[0009] Step 4: Determine the arrangement of the filling balls of the basic corn kernel model based on the analysis results of the geometric morphology;

[0010] Step 5: Based on the geometric parameters of the corn kernel model and the arrangement of the filling balls, the coordinates of the filling balls are calculated to establish the geometric model of the corn kernel;

[0011] Step 6: Determine the contact parameters of the basic model through experiments, predict the changes in surface adhesion behavior through the Sigmoid function combined with the matrix equation and the Linear Cohesion model, and establish a humidity gradient coupled HM-LC model.

[0012] A further technical solution is that in step 1, the corn kernel group selects dent-shaped corn kernels as the representative shape; the filling ball model includes a basic model and a variable model, the basic model is a 12-ball model of corn kernels with a moisture content of 12%, and the variable model is a 12-ball model of corn kernels with a moisture content greater than 12%, and the basic radius and position relationship are determined by the variable position size.

[0013] According to a further technical solution, in step 2, the analysis result of the characteristic size parameters is as follows: the characteristic sizes of the basic corn kernels approximately obey a normal distribution, and there is a correlation between the characteristic sizes;

[0014] The characteristic size parameters include at least the upper base width W1, the lower base width W2, the upper base thickness T1, the lower base thickness T2, the radicle-free height H1, and the total height H2; the correlation of the characteristic size includes the fourth power relationship between the lower base thickness T2 and the upper base width W1, the lower base width W2, the upper base thickness T1, the radicle-free height H1, and the total height H2;

[0015] Taking the Dika C3288 variety as an example, the relationship between different characteristic parameters was determined;

[0016] The fourth power relationship between the bottom thickness T2 and the top width W1 is:

[0017] W1=((-0.0801T2+0.7495)T2) 4

[0018] The fourth power relationship between the bottom base thickness T2 and the bottom base width W2 is:

[0019] W2=((-0.0931T2+0.8420) T2) 4

[0020] The fourth power relationship between the lower base thickness T2 and the upper base thickness T1 is:

[0021] T1=((-0.0648T2+0.5980)T2) 4

[0022] The fourth power relationship between the bottom thickness T2 and the height without radicle H1 is:

[0023] H1=((-0.0500T2+0.57436)T2) 4

[0024] The fourth power relationship between the bottom thickness T2 and the total height H2 is:

[0025] H2=((-0.0594T2+0.6748) T2) 4 .

[0026] In a further technical solution, the geometric parameters of the corn kernel model are simplified to a tetrahedral model, including the upper base width W of the tetrahedral model. q1 , the upper base thickness of the tetrahedron is T q1 , the lower base width of the tetrahedron is W q2 , the bottom base thickness of the tetrahedron is T q2 , the height of the pyramid H q1 ;

[0027] The four vertices of the pyramid are the coordinates of point A ( x A , y A , 0), coordinates of point B (- x A , y A , 0), coordinates of point C (- x A , y C , 0), coordinates of point D (- x A , - y C , 0);

[0028] x A The relationship between the size of the tetrahedron model is:

[0029] ;

[0030] y A The relationship between the size of the tetrahedron model is:

[0031] ;

[0032] y CThe relationship between the size of the tetrahedron model is:

[0033] ;

[0034] X-direction variable coefficient The relationship between the size of the tetrahedron model is:

[0035] ;

[0036] Y-direction variable coefficient The relationship between the size of the tetrahedron model is:

[0037] ;

[0038] Length change value The relationship between the size of the tetrahedron model is:

[0039] ;

[0040] Radius change value The relationship between the size of the tetrahedron model is:

[0041] .

[0042] A further technical solution is to establish a moisture content W in step 3. i With the simulation parameter sequence Y i The quantitative relationship between the deformation moisture content vector of corn kernels is W = [ ω 1, ω 2, ω 3, ω 4, ω 5, ω 6] T Construct the logarithmic feature matrix X:

[0043] ;

[0044] Then, the parameter matrix is ​​constructed, and the simulation parameter sequence Y corresponds to the deformation variable parameter L of the corn. W1 , L W2 , L H1 , L H2 , L T1 , L T2 (L is the size variable value), the simulation parameter sequence Y is obtained as:

[0045] ;

[0046] After that, a regularized least squares solution is performed to solve the prediction matrix C by using the regularized ridge regression method, where the regularization parameter λ = 10 −4, used to suppress ill-conditioned problems and enhance numerical stability;

[0047] ;

[0048] Where I is a unit vector;

[0049] Finally, the prediction parameters are performed. When there is a new water content ω new When first constructing the vector ω vec :

[0050] ;

[0051] The parameter prediction formula for the new moisture content is:

[0052] ;

[0053] in, Represents the predicted Y matrix.

[0054] A further technical solution is that in step 4, the corresponding filling ball position is constructed with XOY as the coordinate plane, with (-( w 1- t 1+1 / 4 L W1 -1 / 4 L T1 ), y A - t 1) Create a circle with a radius of t 1+1 / 4 L T1 The ball O1;

[0055] by(-( w 2- t 2+1 / 4L W1 -1 / 4L T1 ), y C + t 2) Create a circle with a radius of t 2+1 / 4L T1 The ball O2;

[0056] With the Y axis as the axis of symmetry, sphere O1 and sphere O2 are used to form sphere O3 and sphere O4 respectively;

[0057] With (0, y C + t 2) Create a circle with a radius of t 1+1 / 4L H1 ball O5;

[0058] by Create a circle with a radius of ball O6;

[0059] by Create a circle with a radius of The ball O7;

[0060] by Create a circle with a radius of The ball O8;

[0061] With the Y axis as the axis of symmetry, balls O6, O7 and O8 form balls O9 and O 10 With Ball O 11 ;

[0062] With (0, y A ) creates a circle with a radius of 2 as the center h 2-2 h 1+1 / 4L H2 Ball O 12 ;

[0063] In the above formula, w 1. w 2. t 1. t 2. h 1 and h 2 correspond to the actual values ​​of upper base width W1, lower base width W2, upper base thickness T1, lower base thickness T2, height without radicle H1 and total height H2 respectively.

[0064] In a further technical solution, in step 6, the contact model is a humidity gradient coupled HM-LC model, specifically, the Hertz-Mindlin contact model (HM adhesion model) is coupled with the Linear Cohesion contact model (LC adhesion model, i.e., linear adhesion model) to express the surface adhesion form;

[0065] In the LC parameter moisture content, set the moisture content between corn kernels to A, between corn kernels and galvanized steel plate to B, between corn kernels and organic glass plate to C, and between corn kernels and ABS to D, and obtain the simulation parameter sequence Y LC for:

[0066] ,

[0067] Continue to execute the regularized least squares solution and the corresponding prediction function to obtain the corresponding parameter prediction values.

[0068] The Young's modulus of corn kernels is a variable value, and the corresponding equation is obtained by fitting as follows:

[0069] ;

[0070] Where E is Young's modulus, is the moisture content.

[0071] The embodiment of the present invention provides a method for constructing a multi-moisture corn kernel model based on the discrete element method, which has the following beneficial effects:

[0072] (1) By using the measured characteristic size parameters and the filling ball arrangement algorithm, an irregular geometric body based on the real grain morphology (with 12% moisture content grain as the basic model) was constructed. Through the statistical analysis of the characteristic size parameters and the probability distribution modeling, the joint distribution of moisture content, size, and physical property parameters was introduced into the population model.

[0073] (2) A continuous modeling method for the moisture content range of 12%-30% during the entire harvest period was proposed for the first time. The nonlinear change of volume size with moisture content was quantified through Sigmoid function and matrix equation, and a humidity gradient coupled HM-LC model was established to dynamically adjust the surface adhesion parameters. From the five dimensions of multi-moisture content coupling, geometric accuracy, population modeling, parameter prediction efficiency and full-process applicability, the problems of insufficient moisture content coupling, simplified geometric shape, lack of population characteristics and strong experimental dependence in the existing corn kernel modeling process were solved. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 This is a physical picture of corn kernels with different moisture contents;

[0075] Figure 2 This is the size distribution diagram of the characteristic size of corn kernels with a moisture content of 12%;

[0076] Figure 3 This is the partial correlation diagram of the characteristic dimensions of corn kernels with a moisture content of 12%;

[0077] Figure 4 It is a diagram of the geometric model of the tetrahedral structure and corn kernels;

[0078] Figure 5 The discrete element model of corn kernels with different moisture contents;

[0079] Figure 6 Test and simulation images for the calibration of the HM-LC model for corn kernels. DETAILED DESCRIPTION

[0080] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0081] The specific implementation of the present invention is described in detail below with reference to specific embodiments.

[0082] like Figures 1-6 As shown, a method for constructing a multi-moisture corn kernel model based on discrete element method is provided in one embodiment of the present invention, comprising the following steps:

[0083] Step 1: Analyze the geometric morphology and characteristic size parameters of corn kernels at maturity with a moisture content of 12%-30% for the modeled variety, determine the changes in the characteristic size of corn kernels under different moisture contents, and select corn kernels with a moisture content of 12% as the basic model;

[0084] Step 2: According to the analysis results of the characteristic size parameters, the geometric parameters of the basic corn kernel model are obtained;

[0085] Step 3: Use the Sigmoid function with the matrix equation to predict the volume size changes of grains with different moisture contents;

[0086] Step 4: Determine the arrangement of the filling balls of the basic corn kernel model based on the analysis results of the geometric morphology;

[0087] Step 5: Based on the geometric parameters of the corn kernel model and the arrangement of the filling balls, the coordinates of the filling balls are calculated to establish the geometric model of the corn kernel;

[0088] Step 6: Determine the contact parameters of the basic model through experiments. At the same time, the changes in surface adhesion behavior are predicted by combining the Sigmoid function with the matrix equation and the Linear Cohesion model to establish a humidity gradient coupled HM-LC model.

[0089] like Figure 1-Figure 2 As shown, as a preferred embodiment of the present invention, in the step 1, the corn kernel group selects dent corn kernels as the representative shape, and the basic parameters of kernels of other shapes are obtained by analogy according to this method; the filling ball model includes a basic model and a variable model; the basic model is a 12-ball model of corn kernels with a moisture content of 12%; the variable model is a 12-ball model of corn kernels with a moisture content greater than 12%, and the basic radius and position relationship are determined by the variable position size.

[0090] like Figure 3 As shown, as a preferred embodiment of the present invention, in step 2, the analysis result of the characteristic size parameters is: the characteristic size of the basic corn kernel approximately obeys the normal distribution, and there is a correlation between the characteristic sizes;

[0091] The characteristic size parameters include at least the main dimensions such as the upper base width W1, the lower base width W2, the upper base thickness T1, the lower base thickness T2, the height without radicle H1, and the total height H2; the correlation of the characteristic dimensions includes the fourth power relationship between the lower base thickness T2 and the upper base width W1, the lower base width W2, the upper base thickness T1, the height without radicle H1, and the total height H2;

[0092] Taking the Dika C3288 variety as an example, the relationship between different characteristic parameters was determined;

[0093] The fourth power relationship between the bottom thickness T2 and the top width W1 is:

[0094] W1=((-0.0801T2+0.7495)T2) 4

[0095] The fourth power relationship between the bottom base thickness T2 and the bottom base width W2 is:

[0096] W2=((-0.0931T2+0.8420) T2) 4

[0097] The fourth power relationship between the lower base thickness T2 and the upper base thickness T1 is:

[0098] T1=((-0.0648T2+0.5980)T2) 4

[0099] The fourth power relationship between the bottom thickness T2 and the height without radicle H1 is:

[0100] H1=((-0.0500T2+0.57436)T2) 4

[0101] The fourth power relationship between the bottom thickness T2 and the total height H2 is:

[0102] H2=((-0.0594T2+0.6748) T2) 4 .

[0103] like Figure 5 As shown in FIG. 1 , as a preferred embodiment of the present invention, the geometric parameters of the corn kernel model are simplified to a quadrangular pyramid model, including the upper base width W of the quadrangular pyramid model. q1 , the upper base thickness of the tetrahedron is T q1 , the lower base width of the tetrahedron is W q2 , the bottom base thickness of the tetrahedron is T q2 , the height of the pyramid H q1 ;

[0104] The four vertices of the pyramid are the coordinates of point A ( x A , yA , 0), coordinates of point B (- x A , y A , 0), coordinates of point C (- x A , y C , 0), coordinates of point D (- x A , - y C , 0);

[0105] x A The relationship between the size of the tetrahedron model is:

[0106] ;

[0107] y A The relationship between the size of the tetrahedron model is:

[0108] ;

[0109] y C The relationship between the size of the tetrahedron model is:

[0110] ;

[0111] X-direction variable coefficient The relationship between the size of the tetrahedron model is:

[0112] ;

[0113] Y-direction variable coefficient The relationship between the size of the tetrahedron model is:

[0114] ;

[0115] Length change value The relationship between the size of the tetrahedron model is:

[0116] ;

[0117] Radius change value The relationship between the size of the tetrahedron model is:

[0118] .

[0119] As a preferred embodiment of the present invention, in step 3, the Sigmoid function is an S-shaped function commonly found in biology, also known as an S-shaped growth curve. In deep learning, due to its monotonic increasing and inverse monotonic increasing properties, it is often used as an activation function in neural networks to map variables to the range [0, 1].

[0120] Establish moisture content W i With the simulation parameter sequence Y i The quantitative relationship between the deformation moisture content vector of corn kernels is W = [ ω 1, ω 2, ω 3, ω 4, ω 5, ω 6] T Construct the logarithmic feature matrix X:

[0121] ;

[0122] Then, the parameter matrix is ​​constructed, the simulation parameter sequence Y, and the corresponding deformation variable parameter L of the corn is W1 、L W2 、L H1 、L H2 、L T1 、L T2 (L is the size variable value), the simulation parameter sequence Y is obtained as:

[0123] ;

[0124] After that, a regularized least squares solution is performed to solve the prediction matrix C by using the regularized ridge regression method, where the regularization parameter λ = 10 −4 , used to suppress ill-conditioned problems and enhance numerical stability;

[0125] ;

[0126] Where I is a unit vector;

[0127] Finally, the prediction parameters are performed. When there is a new water content ω new When first constructing the vector ω vec :

[0128] ;

[0129] The parameter prediction formula for the new moisture content is:

[0130] ;

[0131] in, Represents the predicted Y matrix.

[0132] like Figure 4 As shown, as a preferred embodiment of the present invention, step 4 includes the following steps:

[0133] Use XOY as the coordinate plane to construct the corresponding filling ball position, (-( w 1- t 1+1 / 4 L W1 -1 / 4 L T1 ), y A - t 1) Create a circle with a radius of t 1+1 / 4 L T1 The ball O1;

[0134] by(-( w 2- t 2+1 / 4L W1 -1 / 4L T1 ), y C + t 2) Create a circle with a radius of t 2+1 / 4L T1 The ball O2;

[0135] With the Y axis as the axis of symmetry, sphere O1 and sphere O2 are used to form sphere O3 and sphere O4 respectively;

[0136] With (0, y C + t 2) Create a circle with a radius of t 1+1 / 4L H1 ball O5;

[0137] by Create a circle with a radius of ball O6;

[0138] by Create a circle with a radius of The ball O7;

[0139] by Create a circle with a radius of The ball O8;

[0140] With the Y axis as the axis of symmetry, balls O6, O7 and O8 form balls O9 and O 10 With Ball O 11 ;

[0141] With (0,y A ) creates a circle with a radius of 2 as the center h 2-2 h 1+1 / 4L H2 Ball O 12 ;

[0142] In the above formula, w 1. w 2. t 1. t 2. h 1 and h 2 correspond to the actual values ​​of upper base width W1, lower base width W2, upper base thickness T1, lower base thickness T2, height without radicle H1 and total height H2 respectively.

[0143] like Figure 5 and 6 As shown in FIG. 1 , as a preferred embodiment of the present invention, in step 6, the contact model is a humidity gradient coupled HM-LC model, specifically, the HM adhesion model is coupled with the LC adhesion model to achieve the expression of the surface adhesion form;

[0144] In the LC parameter moisture content, set the moisture content between corn kernels to A, between corn kernels and galvanized steel plate to B, between corn kernels and organic glass plate to C, and between corn kernels and ABS to D, and obtain the simulation parameter sequence Y LC for:

[0145] ,

[0146] Continue to execute the regularized least squares solution and the corresponding prediction function to obtain the corresponding parameter prediction values.

[0147] The Young's modulus of corn kernels is a variable value, and the corresponding equation is obtained by fitting as follows:

[0148] ;

[0149] Where E is Young's modulus, is the moisture content.

[0150] Other basic parameters are fixed values. For specific values ​​of other parameters, see Table 1 below.

[0151] Table 1

[0152]

[0153] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for constructing a multi-moisture corn kernel model based on discrete element method, characterized in that: The following steps are involved: Step 1: Analyze the geometric morphology and characteristic size parameters of corn kernels at maturity with a moisture content of 12%-30% for the modeled variety, determine the changes in the characteristic size of corn kernels under different moisture contents, and select corn kernels with a moisture content of 12% as the basic model; Step 2: According to the analysis results of the characteristic size parameters, the geometric parameters of the basic corn kernel model are obtained; Step 3: Use the Sigmoid function with the matrix equation to predict the volume size changes of grains with different moisture contents; Step 4: Determine the arrangement of the filling balls of the basic corn kernel model based on the analysis results of the geometric morphology; Step 5: Based on the geometric parameters of the corn kernel model and the arrangement of the filling balls, the coordinates of the filling balls are calculated to establish the geometric model of the corn kernel; Step 6: Determine the contact parameters of the basic model through experiments. At the same time, use the Sigmoid function combined with the matrix equation and the Linear Cohesion model to predict the changes in surface adhesion behavior and establish a humidity gradient coupled HM-LC model. In step 1, a dent-shaped corn kernel is selected as a representative shape of the corn kernel population; the filling ball model includes a basic model and a variable model, wherein the basic model is a 12-ball model of corn kernels with a moisture content of 12%, and the variable model is a 12-ball model of corn kernels with a moisture content greater than 12%. The basic radius and position relationship are determined by the variable position size; In step 2, the analysis result of the characteristic size parameters is: the characteristic size of the basic corn kernels approximately obeys a normal distribution, and there is a correlation between the characteristic sizes; The characteristic size parameters include at least the upper base width W1, the lower base width W2, the upper base thickness T1, the lower base thickness T2, the radicle-free height H1, and the total height H2; the correlation of the characteristic size includes the fourth power relationship between the lower base thickness T2 and the upper base width W1, the lower base width W2, the upper base thickness T1, the radicle-free height H1, and the total height H2; Taking the Dika C3288 variety as an example, the relationship between different characteristic parameters was determined; The fourth power relationship between the bottom thickness T2 and the top width W1 is: W1=((-0.0801T2+0.7495)T2) 4 ; The fourth power relationship between the bottom base thickness T2 and the bottom base width W2 is: <h2 style=";text-align:left;direction:ltr">W2=((-0.0931T2+0.8420) T2)<h2 style=";text-align:left;direction:ltr"> 4 <h2 style=";text-align:left;direction:ltr"> ; The fourth power relationship between the lower base thickness T2 and the upper base thickness T1 is: T1=((-0.0648T2+0.5980)T2) 4 ; The fourth power relationship between the bottom thickness T2 and the height without radicle H1 is: H1=((-0.0500T2+0.57436)T2) 4 ; The fourth power relationship between the bottom thickness T2 and the total height H2 is: <h2 style=";text-align:left;direction:ltr">H2=((-0.0594T2+0.6748) T2)<h2 style=";text-align:left;direction:ltr"> 4 <h2 style=";text-align:left;direction:ltr"> 。 2. The method for constructing a multi-moisture corn kernel model based on discrete element method according to claim 1, characterized in that: The geometric parameters of the corn kernel model are simplified to a tetrahedral model, including the upper base width W of the tetrahedral model. q1 , the upper base thickness of the tetrahedron is T q1 , the lower base width of the tetrahedron is W q2 , the bottom base thickness of the tetrahedron is T q2 , the height of the pyramid H q1 ; The four vertices of the pyramid are the coordinates of point A ( x A , y A , 0), coordinates of point B (- x A , y A , 0), coordinates of point C (- x A , y C , 0), coordinates of point D (- x A , - y C , 0); x A The relationship between the size of the tetrahedron model is: ； y A The relationship between the size of the tetrahedron model is: ； y C The relationship between the size of the tetrahedron model is: ； X-direction variable coefficient The relationship between the size of the tetrahedron model is: ； Y-direction variable coefficient The relationship between the size of the tetrahedron model is: ； Length change value The relationship between the size of the tetrahedron model is: ； Radius change value The relationship between the size of the tetrahedron model is: 。 3. The method for constructing a multi-moisture corn kernel model based on discrete element method according to claim 2, characterized in that: In step 3, the moisture content W is established. i With the simulation parameter sequence Y i The quantitative relationship between the deformation moisture content vector of corn kernels is W = [ ω 1, ω 2, ω 3, ω 4, ω 5, ω 6] T Construct the logarithmic feature matrix X: ； Then, the parameter matrix is ​​constructed, and the simulation parameter sequence Y corresponds to the deformation variable parameter L of the corn. W1 , L W2 , L H1 , L H2 , L T1 , L T2 , L represents the size variable value, and the simulation parameter sequence Y is obtained as: ； After that, a regularized least squares solution is performed to solve the prediction matrix C using the regularized ridge regression method, where the regularization parameter , used to suppress ill-conditioned problems and enhance numerical stability; ； Where I is a unit vector; Finally, the prediction parameters are performed. When there is a new water content ω new When , first construct the vector ω vec : ； The parameter prediction formula for the new moisture content is: ； in, Represents the predicted Y matrix.

4. The method for constructing a multi-moisture corn kernel model based on discrete element method according to claim 3, characterized in that: In step 4, the corresponding filling ball position is constructed with XOY as the coordinate plane, with (-( w 1- t 1+1 / 4 L W1 -1 / 4 L T1 ), y A - t 1) Create a circle with a radius of t 1+1 / 4 L T1 The ball O1; by(-( w 2- t 2+1 / 4L W1 -1 / 4L T1 ), y C + t 2) Create a circle with a radius of t 2+1 / 4L T1 The ball O2; With the Y axis as the axis of symmetry, sphere O1 and sphere O2 are used to form sphere O3 and sphere O4 respectively; With (0, y C + t 2) Create a circle with a radius of t 1+1 / 4L H1 ball O5; by Create a circle with a radius of ball O6; by Create a circle with a radius of The ball O7; by Create a circle with a radius of The ball O8; With the Y axis as the axis of symmetry, balls O6, O7 and O8 form balls O9 and O 10 With Ball O 11 ; With (0, y A ) creates a circle with a radius of 2 as the center h 2-2 h 1+1 / 4L H2 Ball O 12 ; In the above formula, w 1. w 2. t 1. t 2. h 1 and h 2 correspond to the actual values ​​of upper base width W1, lower base width W2, upper base thickness T1, lower base thickness T2, height without radicle H1 and total height H2 respectively.

5. The method for constructing a multi-moisture corn kernel model based on discrete element method according to claim 4, characterized in that: In step 6, the contact model is a humidity gradient coupled HM-LC model, specifically, the HM adhesion model is coupled with the LC adhesion model to express the surface adhesion form; In the LC parameter moisture content, set the space between corn kernels as A, the space between corn kernels and galvanized steel plate as B, the space between corn kernels and organic glass plate as C, and the space between corn kernels and ABS as D to obtain the simulation parameter sequence Y LC for: ； Continue to execute the regularized least squares solution and the corresponding prediction function to obtain the corresponding parameter prediction values.

6. The method for constructing a multi-moisture corn kernel model based on discrete element method according to claim 5, characterized in that: The Young's modulus of corn kernels is a variable value, and the corresponding equation is obtained by fitting as follows: ； in, E is Young's modulus, is the moisture content.

Images