Design and Experiment of Automatic Adjustable Transplanting End-Effector Based on Double-Cam

Abstract

In view of the plugged-out end-effector that can adapt only to a specific size of the tray, the needle spacing and angle of the seedling needle are fixed. In this paper, a new type of plugged-out transplanting end-effector is proposed. The end-effector adopts a double-cam structure to automatically adjust the spacing and angle of the seedling needle, which solves the problem of picking seedlings for different sizes of trays. Firstly, the working principle of 72-hole, 128-hole, and 200-hole trays and a plugged-out end-effector was analyzed. The overall structure of the end-effector was designed. Subsequently, the EDEM software was used to construct the pot seedling model and conduct single-factor simulation experiments to identify the range of factors for the subsequent regression orthogonal experiment. Finally, a tray transplanting test platform was built. With the grasping acceleration, penetration angle, insertion depth, and insertion margin ratio as the test factors and the pot seedling breakage rate as the test evaluation indicators. A four-factor three-level orthogonal regression experiment was conducted to establish a regression model of the seedling breakage rate, and its parameters were optimized. The optimal combination is detailed as follows: a 72-hole tray grasping acceleration of 0.28 m/s², a penetration angle of 13°, an insertion depth of 40 mm, and an insertion margin ratio of 15%; a 128-hole tray grasping acceleration of 0.28 m/s², a penetration angle of 12°, an insertion depth of 36 mm, and an insertion margin ratio of 15%; a 200-hole tray grasping acceleration of 0.28 m/s², a penetration angle of 11°, an insertion depth of 32 mm, and an insertion margin ratio of 10%. Under the optimal combination, the breakage rate of 72 holes reached 2.92%. The breakage rate of 128 holes was stable at 1.76%, while that of 200 holes was stable at 0.68%, which is conducive to the study of a general end-effector. The device developed in this study provides an effective solution to taking and throwing different sizes of cavitation trays, thus providing a practical reference for the study of a generic end-effector.

1. Introduction

In recent years, about 60% of vegetables are transplanted by way of plug trays around the world. This technology not only improves the yield of vegetables but also improves the efficiency of land utilization [1,2,3]. Since the end-effector is the core component of mechanical transplanting devices used for pot seedlings, the design of an end-effector tends to have an immediate impact on the outcome of seedlings [4,5]. In general, an end-effector can be divided into a plugged-out type [6,7,8], a clamping type [9,10,11], and a top clamp combined type [12,13,14] according to how the seedlings are picked up. At present, the sizes commonly used for vegetable hole tray transplantation include 72 holes, 128 holes, and 200 holes. In terms of the plugged-out end-effectors, Hu et al. [15] developed a door-type seeding device that could be used for the 72-hole and 128-hole trays. Jiang et al. [16] proposed an efficient end-effector driven by a linear cylinder for a 72-hole tray. Zhao et al. [17] developed an end-effector driven by the cylinder to achieve full cam extraction. During high-speed seedling pick-up and release, the plugged-out end-effectors exhibit significant vibration and poor stability. In terms of the clamping type end-effector, Ye et al. [18] put forward a planetary rotary clamping mechanism for the 128-hole tray. Vivek P [19] designed an efficient mechanical linkage pickup device for a 98-hole tray. Ren et al. [20] developed a stretchable delivery device for the 128-hole tray. The clamping type end-effector exerts a clamping force when picking up the seedlings, but it is difficult to control this clamping force. Also, it is prone to crushing the bowl, which reduces the outcome of transplanting. When the seedlings are sowed, they may adhere to the area of contact, which makes it difficult for them to fall off automatically. In terms of the top clamp combined type end-effector, Wang et al. [21] developed a crank-swing type pot seedling-taking device for the 128-hole tray. Jin et al. [22] proposed an automatic single-row transplanting end-effector for the 128-hole tray. The end-effector fitted with a top clamp combines the advantages of these two methods. The height coordination between different parts is required when the seedlings are picked, the requirements on positioning control for the related working parts are demanding, and the structure is complex. When the top seedling rod reaches a greater height than the pot seedling, it will damage the pot body and affect the outcome of transplanting if the top force is too large. To sum up, there have now been plenty of excellent achievements made in the above studies on vegetable transplanting by hole tray. Despite this, there remain some problems, such as the high breakage rate of pot seedlings and poor versatility.

Vegetable planting in China is characterized by small planting areas, numerous varieties, and widespread distribution. As a result, there are different specifications of vegetable seedlings required for factory hole tray breeding, which drives the demand for end-effectors suited to hole trays of different sizes. When the spacing of the needle and the angle of the end-effector are constant, the breakage rate of pot seedlings increases, and the yield of pot seedlings diminishes. Due to the advantages of the plug-in end-effector, such as structural simplicity, high efficiency, excellent operability, and low cost, the seedlings cause minimal damage to the matrix. In this paper, a new automatic adjustable pluggable vegetable end-effector was proposed that is suitable for 72 holes, 128 holes, and 200 holes trays, which solves the problem of picking seedlings from 72 holes, 128 holes, and 200 holes. By analyzing three types of hole trays, a double-cam structure was developed to adjust the needle distance and angle, which solves the problem with three kinds of hole trays for taking seedlings. Secondly, EDEM simulation was performed, and the key influencing factors in the performance of the end-effector were identified. Finally, a prototype was designed and used for transplanting experiments of pot seedlings, with its performance verified through analysis. Thus, the low versatility of end-effectors was addressed.

2. Materials and Methods

2.1. Establishment of Theoretical Model of Plugged-Out End-Effector Motion

The seedling-picking mechanism of the plugged-out end-effector is shown in Figure 1. Specifically, the prototype of the plugged-out end-effector is shown in Figure 1a. Driven by the cylinder, the seedling needle is stuck into the pot seedling. As the pot seedling is extracted upward by the seedling needle to separate the pot seedling from the plug tray, it moves with the seedling needle to complete seedling pick-up [23]. The penetration angle and the distance from the insertion point to the hole edge, as obtained when the seedling needle is stuck into the hole plate of different sizes, are shown in Figure 1b. D represents the distance from the insertion point to the hole side, φ indicates the penetration angle of the seeding needle, and the penetration angle is defined as the included angle between the seedling needle and the vertical direction. The working principle of the plugged-out end-effector is shown in Figure 1c. Through a literature review and single-factor pre-experiment of seedling needle penetration angle, the following conclusions are drawn [24,25,26]. Firstly, in order to prevent the pot seedling from breaking in the course of taking seedlings, the needle must be stuck into the pot near the inner wall of the hole. If D is too large, it would lead to the overly low volume of the pot seedling wrapped by the four seedling needles within the force range, excessive stress concentration, and severe damage to the pot seedling. Conversely, if D is too small, it would make the seedling needle unduly close to the edge of the pot seedling and would be prone to damaging the edge area and causing the particle to fall off. Secondly, the size and depth of holes vary, which makes it necessary to consider the penetration angle φ of the seedling needle. If φ is too large, the seedling pot tends to suffer significant deformation and serious damage. Conversely, if φ is too small, it would result in the insufficient grasping force of the seedling needle, which causes severe damage to the pot seedling in the vertical direction. Lastly, peat is the main component of the seedling pot, and the cohesive force is limited. Therefore, attention should be paid to prevent the pot from rupturing during seedling pick-up.

2.2. Overall Design of End-Effector and EDEM Simulation

2.2.1. Overall Structure of End-Effector

Figure 2 shows the structure of the end-effector suitable for various sizes of the orifice. The end-effector consists mainly of a stepping motor, frame, coupling, double semicircle cam, slider, cylindrical cam, cylinder, seedling pusher, seedling needle, etc. The double-cam rotation structure is comprised of a double semicircular cam, a cylindrical cam, and a stepper motor. Firstly, the plug seedlings are placed on the conveyor belt. Driven by the conveyor belt, the plug seedlings first pass through the identification mechanism. Then, the specifications of the plug seedlings are determined by the identification mechanism, and the information is fed back to the end-effector. Finally, the relevant working parameters needed to pick up the seedlings are adjusted by the end-effector. Responsible for the adjustment of needle spacing and penetration angle, the end-effector drives the double-cam structure to rotate through the stepper motor. The rotation of the double semicircle cam drives the reciprocating linear movement of the slider along the optical axis, while the rotation of the cylindrical cam causes the H-shaped fork frame to rise or fall, thus changing the spacing of the needle and penetration angle, as can be seen from Figure 2b. The penetration angle is expressed as:

$$\phi =\arcsin \frac{2L_1-a+2D}{2L_5}=\arcsin \frac{2L_1-a+2D}{2\left(L_4+Q+L_3\right)}$$

(1)

where L₁ refers to the distance from the bottom of the cylinder to the center line; L₃ is the length of the seedling needle.; L₄ is the length of the cylinder; L₅ represents the distance from the needle insertion point to the bottom of the cylinder; Q is the stroke of the cylinder; a is the upper caliber side length on the tray.

2.2.2. Key Component Design

In this study, four seedling needles are adopted to take pot seedlings, and the way of picking seedlings is plugged-out.

Table 1. Three sizes hole tray hole size.

Hole Size	Tray Specification/Hole
	72	128	200
Upper caliber side length a/mm	40	32	24
Lower caliber side length b/mm	20	13	10
Height h/mm	45	42	36

lists the hole sizes of the three types of holes. The schematic diagram of the end-effector is shown in Figure 2b.

• Penetration angle φ

The deeper the seedling needle is stuck, the smaller the force of the pot seedling applied by the plug, and the easier it is to take the seedling. The seedling needle is stuck as deep as possible into the pot seedling [27,28]. When the penetration angle is too large, the seedling needle causes interference, thus resulting in damage to the pot seedling. Conversely, when the penetration angle is too small, it is difficult to extract pot seedlings. Therefore:

$$\phi \le {\phi }_{\max }=\arctan \frac{a}{2h}$$

(2)

By calculating the maximum penetration angle of the three kinds of trays, it can be seen that the maximum penetration angle of 200 holes is the smallest, and the maximum penetration angle is 18.43°. The maximum penetration angle of the double-cam adjustment seedling needle is 18°.

2.

Seedling needle insertion length L₂

When the length of the seedling needle is small, the matrix at the bottom of the pot seedling is broken, which increases the damage rate of the matrix and affects the outcome of seedling pick-up. If the seedling needle comes into contact with the bottom of the plug tray, it may damage the plug tray and the dense root system at the bottom. The length of the seedling needle into the soil is defined as the length of the seedling needle inserted into the pot seedling. Therefore:

$$L_2\le \frac{h}{\cos {\phi }_{\max }}=L_{\max }=\frac{\sqrt{a^2+4h^2}}{2}$$

(3)

3.

Seedling needle length L₃

When the length of the seedling needle is smaller than the maximum depth at which the seedling needle is stuck, the matrix at the bottom of the seedling is broken, which affects the outcome of seedling pick-up. In order to prevent the seedling from being damaged during seedling pick-up, the length of the seedling needle must be greater than the maximum depth at which the seedling needle is stuck. The length of the seedling needle should meet the following requirement:

$$L_3\ge L_2$$

(4)

The maximum depth of 72 holes is 45 mm. Since the insertion length of the seedling needle is the largest, the Seedling needle length is 55 mm.

4.

Cylinder stroke Q

The cylinder drives the seedling needle to be stuck into the pot seedling. The stroke of the cylinder determines the depth at which the seedling needle is stuck into the soil, which in turn affects the outcome of the pot seedling. Therefore, the stroke of the cylinder is required to meet the maximum depth at which the seedling needle is stuck during operation. Therefore:

$$Q\ge L_2$$

(5)

The maximum depth of 72 holes is 45 mm. The insertion length of the seedling needle is the largest. Therefore, the cylinder stroke is 60 mm. A 160 mm long cylinder produced by Yueqing Hebai Pneumatic Plant with a 60 mm stroke was used.

5.

Double-cam structure design

In the present study, the double-cam structure is applied to adjust the pitch of the needle and the penetration angle of the end-effector for the collection of plug seedlings of three different sizes. In the single-factor test on penetration angle, the seedling is chosen to be the plug-pulling type when the penetration angle of the seedling needle is less than 9°. Thus, the pot seedling cannot be taken out. Therefore, the penetration angle of the double-cam structure adjustment seedling needle ranges from 9° to 18°. Since the 200-hole plug has the smallest size, the relevant parameters of the double-cam structure can be obtained by analyzing the 200-hole plug. The seedling needle is stuck into the pot seedling at the maximum penetration angle. When the cylinder is fully extended, the tip of the seedling needle comes into contact with the bottom of the plug. At this time, the distance from the bottom of the cylinder to the center is 86.7 mm. When the double-cam structure rotates 180°, the stroke of the slider is 40 mm, the penetration angle of the seedling needle is merely 9°, and the length of the connecting piece is 60 mm. By adjusting the distance between the upper surface of the plug and the tip of the needle, it can be applied to both 72-hole and 128-hole seedlings when the penetration angle is in the range of 9–18°.

Since the double-cam structure rotates at a constant speed, the pressure angle of the cylindrical cam remains unchanged. The design of the double semicircular cam structure is illustrated in Figure 3a, and that of the cylindrical cam structure is shown in Figure 3b.

Slider stroke S:

$$S=\left(R_1-R_3\right)\frac{\omega }{\pi }\left(0\le \omega \le \pi \right)$$

(6)

Radius of base circle of double semicircle cam R₄:

$$R_4=R_3-R_5$$

(7)

Double semicircle cam eccentricity D:

$$D=R_1-R_2$$

(8)

Cylindrical cam lift H₁:

$$H_1=\frac{\omega h_2}{2\pi }\left(0\le \omega \le \pi \right)$$

(9)

In the formula, R₁ is the radius of the seedling needle (86.7 mm) when the penetration angle is 18°; R₂ is the track radius of the double semicircular cam (66.7 mm); R₃ was 46.7 mm when the penetration angle was 9°. R₄ represents the base circle radius of the double semicircular cam (44.2 mm); R₅ is a double semicircular cam chute radius of 2.5 mm; r_m is the average radius of the cylindrical cam (14 mm); D is eccentricity (20 mm); β is the pressure angle of the double semicircular cam (12.68°); γ represents the pressure angle of the cylindrical cam (22.53°); h₂ indicates that the pitch of cylindrical cam is 42 mm.

2.3. Design of Simulation Experiment

Given the tray size of 72 holes, 128 holes, and 200 holes, a three-dimensional model was established and imputed into the EDEM software. The relevant simulation parameters are listed in

Table 2. Structural characteristic parameters of simulation object.

Parameter	Material
	Matrix	Tray	Steel Needle
Poisson ratio	0.4	0.38	0.3
Shear modulus/MPa	10	866	7000
Density/(kg·m−3)	750	1400	7800

. Peat is the main component of pot seedlings. The particle with a radius of 1 mm was set as a single-sphere model [29]. When the seedlings were taken, the bond between the particles broke down, thus causing the particles to fall off. The contact model Hertz-Mindlin with bonding in EDEM was applied to simulate the process of crushing and fracture. The tray was made of polystyrene, and the seedling needle was made of 45 steel. When pot seedlings were extracted, there were different factors that may cause the shedding of pot seedling particles to different degrees. If the particles fall off unduly, it would affect the outcome of seedling extraction. Before the simulation, the most significant factor affecting the damage rate of the pot seedling matrix during transplanting was determined through a literature review and the single-factor experiment on the extraction of pot seedlings. They are grasping acceleration, penetration angle, insertion depth, and insertion margin ratio, respectively. Grasping acceleration refers to the acceleration that takes place when the end-effector is used to extract the pot seedling upward. The insertion margin ratio is defined as the percentage of the distance between the insertion point and the edge line on the hole and the side length of the hole. A single-factor simulation test was carried out to verify the outcome of pot seedling extraction. As shown in Figure 4, the extraction process is accompanied by the shedding of particles. The breakage rate of the pot seedling model was calculated, and the extracted particles were counted by Grid Bin Group in the EDEM post-processing module. Each group was tested 10 times, with the average value taken. The optimal range of each factor was determined by using the seedling breakage rate as the test Factor. The formula used to calculate the pot seedling breakage rate is expressed as:

$$Q=\frac{M-M_1}{M}$$

(10)

where Q represents the damage rate of the seedling substrate (%), M indicates the average weight of the seedling before removal (g), and M₁ denotes the average weight of the seedling after removal (g).

2.4. Transplanting Test Design

In this study, the test was performed on a self-made test platform, as shown in Figure 5, with grab acceleration, penetration angle, insertion depth ratio, and insertion margin ratio as the experimental factors. The pot seedlings on 72-hole, 128-hole, and 200-hole trays with 30 days of seedling age were taken as the research object. The four-factor three-level orthogonal regression test was conducted with the seedling pot breakage rate as the test Factor. Pu chun YP202A electronic balance (accuracy: 0.01 g) was used for weighing, with the average taken. The formula used to calculate Q₁ is expressed as:

$$Q_1=\frac{m_1-m_{{}_2}}{m_1}\times 100\%$$

(11)

where Q₁ is the seedling breakage rate (%); m₁ is the total mass of pot seedling before taking out (g); m₂ is the quality of pot seedling after taking out (g).

2.5. Parameter Optimization and Experimental Verification Design

In order to improve the performance of the end-effector, the working parameters and structural parameters of the end-effector were optimized for the low breakage rate. The mathematical model was established with the minimum breakage rate y as the objective function, the grasping acceleration x₁, the penetration angle x₂, the insertion depth x₃, and the insertion margin ratio x₄ as the constraints. After optimization, the end-effector was verified.

3. Results

3.1. Simulation Result

By changing one of the factors to calculate the breakage rate of the pot seedling, the impact of each factor on the breakage rate was determined. The test results are shown in Figure 6.

• Effect of grasping acceleration on the breakage rate of pot seedling

When the pot seedling is extracted upward by the seedling needle, the instantaneous grasping acceleration changes in the vertical direction were simulated, the results of which are shown in Figure 6a. When the instantaneous grasping acceleration falls below 0.3 m/s², the breakage rate of pot seedlings decreases gradually. This is largely attributed to the limited instantaneous force of grasping acceleration on pot seedlings and the low breakage rate of pot seedlings. When the grabbing acceleration exceeds 0.3 m/s², the breakage rate of pot seedlings increases progressively. This is because the greater the impact force on the pot seedlings, the more significant the matrix shedding. Therefore, the range of grasping acceleration for 72 holes, 128 holes, and 200 holes is 0.2–0.4 m/s².

2.

Effect of penetration angle on breakage rate of pot seedling

The simulation was carried out by adjusting the penetration angle, the results of which are shown in Figure 6b. When the penetration angle is too large, interference occurs, which causes serious damage to the pot seedlings. Conversely, when the penetration angle is too small, the force required for seedling extraction is insufficient for the pot seedling, and the pot body moves downward during seedling pick-up. As a result, the pot seedling is severely damaged in the vertical direction. Therefore, the range of penetration angle for the 72-hole tray is 12–14°, the range of penetration angle for the 128-hole tray is 11–13°, and the range of penetration angle for the 200-hole tray is 10–12°.

3.

Effect of insertion depth on seedling breakage rate

The test results were obtained by simulating the insertion depth, as shown in Figure 6c. Given a large insertion depth, the extrusion force between it and the soil exceeds the cohesion between the particles, thus tearing the soil block. When the insertion depth is limited, the pulling force exerted by the seedling needle is insufficient, thus causing inability when the pot seedling is taken out or broken in the middle, with most of the growth material left in the hole. Therefore, the range of insertion depth is 36–44 mm, 32–40 mm, and 28–36 mm for the 72-hole, 128-hole, and 200-hole trays, respectively.

4.

Effect of insertion margin ratio on seedling breakage rate

By simulating the insertion margin ratio, the test results were obtained, as shown in Figure 6d. A relatively large insertion margin leads to an overly small volume of the pot seedling particles within the force range wrapped by the four seedling needles, an excessive stress concentration, and severe damage. When it is too small, it is easy to damage the edge area of the pot seedling, causing the breakage rate to increase. Therefore, the range of the insertion margin ratio for 72-hole and 128-hole trays is 10–20%, while that for 200-hole trays is 5–15%.

3.2. Results of Transplanting Test

The optimal range of each factor was finalized by means of simulation. Under the same conditions, the prototype developed in this study was used to test the outcome of seedling pick-up for three types of plug seedlings. The experiment was performed in the laboratory of Henan University of Science and Technology in December 2022. The samples were collected from ‘Luo Jiao 308′ cultivated by Luoyang Chengyan Pepper Research Institute. The substrate and tray were provided by the institute, and the main component of the substrate was peat. The moisture content of the pot seedlings used in the experiment ranges between 50% and 60%. For each size of the tray, three groups were tested, each group of 60. The test factors and levels of the orthogonal regression test are shown in

Table 3. Levels of test factors for three trays of different sizes.

Tray Specification/Hole	Level	Factor
		Grasping Acceleration m/s2	Penetration Angle/°	Insertion Depth/mm	Insertion Margin Ratio/%
72	−1	0.2	12	36	10
	0	0.3	13	40	15
	1	0.4	14	44	20
128	−1	0.2	11	32	10
	0	0.3	12	36	15
	1	0.4	13	40	20
200	−1	0.2	10	28	5
	0	0.3	11	32	10
	1	0.4	12	36	15

, while the test results are shown in

Table 4. Test design and results.

Serial Number	Factor				Breakage Rate (%)
	x1	x2	x3	x4	y72	y128	y200
1	0	0	0	0	2.97	2.33	1.33
2	1	−1	0	0	12.18	10.28	9.19
3	0	1	1	0	11.25	9.75	8.75
4	−1	0	0	1	9.32	7.82	6.98
5	1	1	0	0	15.12	13.32	11.83
6	0	−1	1	0	8.28	6.48	5.53
7	0	0	1	−1	8.12	7.13	5.82
8	0	0	0	0	3.15	1.72	1.28
9	0	−1	0	1	6.71	5.21	4.17
10	1	0	0	1	14.02	11.79	10.84
11	0	1	0	1	10.63	8.93	7.63
12	0	0	1	1	9.77	7.13	6.53
13	1	0	−1	0	12.23	11.13	10.03
14	0	0	−1	−1	8.09	6.49	5.47
15	0	1	−1	0	10.41	8.71	8.76
16	−1	1	0	0	11.32	9.82	8.73
17	−1	−1	0	0	8.32	6.82	5.88
18	0	−1	−1	0	6.47	5.17	4.47
19	0	0	0	0	3.58	1.97	0.97
20	−1	0	0	−1	7.11	6.21	5.21
21	0	0	0	0	3.07	2.17	0.72
22	1	0	0	−1	11.86	10.26	9.28
23	1	0	1	0	14.89	12.59	11.95
24	0	0	0	0	3.77	2.48	1.17
25	0	−1	0	−1	8.56	5.23	4.23
26	0	1	0	−1	8.98	8.22	7.19
27	−1	0	1	0	9.75	8.15	7.15
28	−1	0	−1	0	8.49	6.89	5.92
29	0	0	−1	1	7.89	6.73	5.73

, where x₁, x₂, x₃, and x₄ represent the encoding values of grasping acceleration, penetration angle, insertion depth, and insertion margin ratio, respectively.

The quadratic polynomial regression models of grasping acceleration, penetration angle, insertion depth, insertion margin ratio, and pot seedling breakage rate were established respectively. The regression equations are as follows:

$$\begin{array}{cc}\hfill y_{72}& =3.31+2.17x_1+1.43x_2+0.71x_3+0.47x_4-0.015x_1{x}_2+0.35x_1{x}_3-0.013x_1{x}_4-0.24x_2{x}_3+\hfill \\ & 0.88x_2{x}_4+0.46x_3{x}_4+5.18x_{{}_1}^2+3.13x_2^2+2.81x_3^2+2.24x_4^2\hfill \end{array}$$

(12)

$$\begin{array}{cc}\hfill y_{128}& =2.13+1.97x_1+1.63x_2+0.49x_3+0.36x_4+0.01x_1{x}_2+0.05x_1{x}_3-0.02x_1{x}_4-0.068x_2{x}_3+\hfill \\ & 0.18x_2{x}_4+0.0075x_3{x}_4+4.98x_{{}_1}^2+2.84x_2^2+2.62x_3^2+1.97x_4^2\hfill \end{array}$$

(13)

$$\begin{array}{cc}\hfill y_{200}& =1.09+1.94x_1+1.62x_2+0.45x_3+0.39x_4-0.052x_1{x}_2+0.17x_1{x}_3-0.053x_1{x}_4-0.27x_2{x}_3+\hfill \\ & 0.12x_2{x}_4+0.11x_3{x}_4+4.94x_{{}_1}^2+2.86x_2^2+2.83x_3^2+1.95x_4^2\hfill \end{array}$$

(14)

According to

Table 5. Regression model analysis of variance.

Factors	Variance Source	Sum of Squares	Degree of Freedom	Mean Square	F Value	p Value
y72	Model	318.40	14	22.74	60.88	<0.0001
	Residual error	5.23	14	0.37
	Lack of fit	4.75	10	0.47	3.93	0.10
	Error	0.48	4	0.12
y128	Model	283.34	14	20.24	110.63	<0.0001
	Residual error	2.56	14	0.18
	Lack of fit	2.20	10	0.22	2.46	0.2
	Error	0.36	4	0.089
y200	Model	283.67	14	20.26	78.20	<0.0001
	Residual error	3.63	14	0.26
	Lack of fit	3.38	10	0.34	5.37	0.06
	Error	0.25	4	0.063

Note: p < 0.01 means extremely significant.

, the regression equation model of the breakage rate is p < 0.0001, indicating the significance of the regression equation model. It is also suggested that the regression model fits well within the test range.

The Influence of the Interaction of Various Factors on the Breakage Rate

To find out how interaction factors affect the seedling breakage rate, a 200-hole tray was used as an example to analyze the impact of the interaction on the breakage rate of pot seedlings by changing two factors. The response surface is shown in Figure 7.

As shown in Figure 7a, the grabbing acceleration and penetration angle increase. The breakage rate first decreases and then increases. When the grabbing acceleration gets close to 0.3 m/s² and the penetration angle reaches about 11°, the breakage rate is low, as shown in Figure 7b. When the grasping acceleration is constant, the breakage rate decreases first and then increases with the rise of insertion depth. When the insertion depth is constant, the breakage rate declines first and then rises with the increase of grasping acceleration. When the grasping acceleration reaches about 0.3 m/s² and the insertion depth gets close to 32 mm, the breakage rate is low, as can be seen in Figure 7c. When the grasping acceleration is constant, the breakage rate decreases first and then increases with the rise of the insertion margin ratio. When the insertion margin ratio is constant, the breakage rate decreases first and then increases with the rise of grasping acceleration. When the grasping acceleration reaches about 0.3 m/s² and the insertion margin ratio gets close to 10%, the breakage rate is low, as shown in Figure 7d. When there is no change in one of the factors for the penetration angle and the insertion depth, with another factor increasing, the breakage rate decreases first and then increases. When the penetration angle reaches about 11° and the insertion depth reaches about 32 mm, the breakage rate is low, as can be seen in Figure 7e. When there is no change in one of the factors for the penetration angle and the insertion margin ratio, the other factor increases, which causes the breakage rate to decrease first and then increase. When the penetration angle reaches about 11° and the insertion margin ratio reaches about 10%, the breakage rate is low, as shown in Figure 7f. When there is an increase in any factor for the insertion depth and insertion margin ratio, the breakage rate decreases first and then increases. The breakage rate is low when the insertion depth is about 32 mm and the insertion margin ratio is about 10%.

3.3. Optimization and Verification Results

The optimal combination, as determined by Formula (14), is shown in

Table 6. Working parameters and pot seedling breakage rate before and after optimization.

	Tray Size/Hole	Grasping Acceleration/m2	Penetration Angle/°	Insertion Depth/mm	Insertion Margin Ratio/%	Breakage Rate/%
Before optimization	72	0.28	12.77	39.53	14.76	2.87
	128	0.28	11.72	35.62	14.60	1.67
	200	0.28	10.71	31.66	9.54	0.64
After optimization	72	0.28	13	40	15	2.92
	128	0.28	12	36	15	1.76
	200	0.28	11	32	10	0.68

. In order to verify the outcome of optimization and ensure the accuracy of control on the end-effector, the working parameters were set and the double-cam structure was verified through experiment. Each group was repeated 10 times on the seedling extraction test bench, with the average taken. Table 6 lists the optimized parameter combination and verification results. In order to ensure the smooth operation of the double-cam structure, it is necessary to ensure that the maximum pressure angle of the mechanism is less than 30° [30]. The parameters related to the optimized double-cam structure are shown in

Table 7. Double-cam structure parameters after optimization.

R1	R2	R3	R4	R5	D	S	rm	h2	βmax	γmax
64.8 mm	58.5	52.3 mm	49.8 mm	2.5 mm	6.3 mm	12.4 mm	14 mm	12 mm	6.13°	7.77°

. The effect of seedling extraction is shown in Figure 8. By comparing the results, it can be found that the results of theoretical optimization are basically consistent with the experimental results, which confirms that the regression model is reliable.

4. Discussion

The outcome of seedling pick-up achieved by the three trays after analysis is satisfactory. The results show that the optimal value obtained by theoretical calculation is very close to the experimental result and that the reliability of the regression model is sufficient, which confirms the accuracy of the simulation model construction. The shape of the matrix particles set in the simulation is spherical, and the change in particle size determines the number of particles generated. Differently, the change in soil particle size in the simulation has little effect on the simulation results, so it is considered negligible. Through comparison with the existing universal end-effector, it was found that the end-effector developed in this paper has two main advantages. On the one hand, it can be used to prevent the seedling from damage and breaking when the end-effector works, thereby reducing the risk of seedling injury and the breakage rate of pot seedlings. On the other hand, it can be used to adjust the distance between the fingers and the entry angle of the seedling needle, which makes it suitable for different sizes of trays. Also, the breakage rate of pot seedlings is reduced to a very low level.

There are several reasons for the high breakage rate of pot seedlings. Firstly, under the same conditions of cultivation, the sizes of 128-hole and 200-hole trays are smaller than those of the 72-hole trays, which makes it easy for the roots to be formed when pot seedlings better wrap particles. Thus, in the process of transplanting, the pot seedling particles are scattered less, and the damage is relatively limited. The moisture content of the substrate can increase the cohesive force inside the pot seedlings, or that between the matrix particles, to be precise. Also, it is easier to extract the pot seedlings. The water content of the pot seedlings selected for this prototype test is in the range of 50–60%, while it still varies between the pot seedlings. The effect on the damage rate of the pot seedling matrix remains. At the same time, it was found in the actual transplanting process that damage to pot seedlings caused by the end-effector during seedling pick-up resulted from the extrusion of the seedling needle on the pot seedling particles during the insertion process, which leads to the mutual changes between particles as well as between particles and roots, thus causing the deformation of pot seedlings. As a result, the particles on the upper surface of the pot seedling expand outward and fall off. Secondly, in the process of transplanting, the inaccurate positioning of the trays and the positioning error of the mechanical arm cause the insufficient insertion depth of the seedling needle, and the pot seedling cannot be completely taken out, thus damaging the pot seedling particles severely and reducing the outcome of seedling extraction. At the same time, the pot seedling particles are loose due to the poor growth of some pot seedling roots, and the pot seedling particles fall off in large amounts during seedling extraction, thus resulting in the failure of transplanting.

5. Conclusions

(1) To address the poor versatility of the plugged-out end-effector, its application to trays of different sizes should be avoided. A new type of automatic adjustable seedling extraction end-effector was developed in this study. The double-cam structure was composed of a double semicircle cam and a cylindrical cam driven by a stepper motor to adjust the needle distance and the penetration angle. Under the action of the cylinder, the seedling needle is tilted into the pot seedling, which reduces the damage caused by the end-effector to the pot seedling. Finally, the critical parameters of each institution are identified.

(2) EDEM simulation software was applied to carry out a single-factor simulation test on grasping acceleration, penetration angle, insertion depth, and insertion margin ratio, with the test range of relevant parameters determined. Then, the prototype test was carried out. The test data were analyzed through a four-factor, three-level orthogonal regression test. According to the test results, grasping acceleration, penetration angle, insertion depth, and insertion margin ratio have a significant impact on the breakage rate of the pot seedling matrix. The optimal combination is finalized as follows: a grasping acceleration of 0.28 m/s², a penetration angle of 13°, an insertion depth of 40 mm, an insertion margin ratio of 15% for the 72-hole tray; a grasping acceleration of 0.28 m/s², a penetration angle of 12°, an insertion depth of 36 mm, and an insertion margin ratio of 15% for the 128-hole tray; a grasping acceleration of 0.28 m/s², a penetration angle of 11°, an insertion depth of 32 mm, and an insertion margin ratio of 10% for the 200-hole tray. Under the optimal combination, the breakage rate of the 72-hole pot seedlings was 2.92%, that of the 128-hole pot seedlings was 1.76%, and that of the 200-hole pot seedlings was 0.68%. This provides guidance for the study of universal end-effectors.

(3) In this paper, an innovative end-effector was proposed that can be applied to cavity trays of different sizes. By changing the parameters of the double-cam structure, the operation of cavity trays of different sizes was achieved.

(4) The end-effector was developed mainly to solve the problem with its kinematics force, with the root system and water content of pot seedlings ignored. This will be further improved in the future.