Bionic Optimization Design and Experiment of Reciprocating Cutting System on Single-Row Tea Harvester

Abstract

The reciprocating cutting system is one of the key parts of a tea harvester; and its cutting performance directly determines the cutting power consumption and harvesting quality of the machine. Because the structural parameters of reciprocating cutting systems do not match the tea cut; resulting in larger cutting resistance, it is necessary to optimize the structural parameters. The cricket mouth part has outstanding performance in tea tree fiber cutting; and the curved structural characteristics of the upper jaw of the cricket have been useful to improve the cutting efficiency of cutting system. Quantitative analysis of the structure of the upper jaw revealed that the arc-shaped structure of the incisor lobe would inspire new bionic blades and bionic cutters to solve the above problems. The cutting performance experiment of the cutting blade was designed for investigating the effects of inter-node number; tea variety and blade type (ordinary blade; bionic blade e and bionic blade f) on the cutting force and cutting power consumption. Experimental results of cutting performance have shown that the bionic blade could reduce cutting resistance and cutting power consumption. Tea varieties had little effect on cutting force and cutting power consumption. In addition, the orthogonal test was carried out to study the influence of cutter type with the cutting speed ratio and cutting angle on the integrity rate and missing rate of tea shoot. The field cutting experiment showed that the cutting angle was the most important for the integrity rate and missing rate of tea shoot; followed by the cutter speed ratio; and finally, the cutter type. The optimum combination of parameters was a cutting speed of 0.8 m/s; a forward speed of 1.0 m/s; a cutting angle of −3°, and using the bionic cutter e. With the optimal parameter combination, the integrity rate and missing rate of the tea shoot were 92.7% and 3.9%, which were increased by 13.2% and decreased by 6.4% compared to those under the condition of the 0° cutting angle and an ordinary cutter. As a result, the bionic cutter could obviously reduce cutting resistance; reduce cutting power consumption and improve the harvesting quality. These results would provide guidance for the design of the reciprocating cutting system of tea harvesters and other stalk cutting machinery.

1. Introduction

Tea is one of the most popular health beverages in the world [1,2]. Over 30 countries and regions around the globe cultivate and produce tea. In China, there are about 3.26 million hectares of tea planting area, and annual output of dry tea is over 3.06 million tons, accounting for over 60% and 45% of the world’s tea planting area and production, respectively, ranking first in the world [3,4]. The tea industry has always been a significant economic partner for China. In many areas of tea production, the cost of the harvest is the primary constraint in improving the economic benefits [5,6]. In order to reduce the cost of the harvest, it is necessary to use a tea harvester instead of harvesting it manually. Nowadays, tea harvesters are widely used for the mechanized harvest of tea [7,8,9]. However, the used harvester has a series of drawbacks, such as low harvesting quality and high cutting power consumption.

The cutter is the main working part of the tea harvester. The shape and parameters of the cutter have a great influence on the quality of the harvest and cutting power consumption of the tea harvester [10,11]. A great deal of research has been carried out on the shape and structure of the cutter to reduce cutting power consumption and improve harvesting quality [12,13,14]. The foreign tea harvester has had a long history of development and harvesting methods have been generally divided into three types: break harvest, draw harvest and cut harvest. When the break harvest method was used, there were many intact tea shoots, and the missing rate of the tea shoot was as high as 25~40%. This harvesting method causes greater damage to tea trees. The required working parts by this method have a complex structure and low work efficiency, so it is less applied. The draw harvest method was in the primary experiment stage, which had not yet been applied. When the cut harvest method is adopted, the required working parts have the advantages of simple structure, high work efficiency, low missing rate, and little damage to tea trees. In order to meet the requirements of high work efficiency, combined with the actual production demand, the commonly used harvesting method is cut harvest with a reciprocating cutter [15]. Liu et al. used ADAMS software to establish the kinematic simulation model of a reciprocating cutter to analyze and optimize the cutter speed parameters [16]. Song et al. established the three-dimensional model of a tea harvester, which included a walking organization, transmission system, elevator organization and reciprocating cutting system, and verified the harvesting quality of the reciprocating cutting system [17]. Wu et al. developed a single tea harvester and studied the influence of the structural parameters of the reciprocating cutter on cutting performance. It was determined that the missing cutting rate to the tea shoot was 2% and the integrity rate was 85% when the cutter structural parameter of the cutting width was 600 mm [18].

In the harvesting process of the crop, the cutting resistance and cutting power consumption are important indexes to evaluate the cutting performance. In recent years, bionic technology has been widely used to optimize the structural parameters of the cutter for a reduction in cutting resistance and cutting power consumption [19,20,21,22]. Yang et al. designed a bionic soil-cutting blade based on the multi-claw combination of mole rat. The rake angle and soil moisture content had significant effects on cutting resistance [23]. Tong and Xu discovered that the contour curve of the incisor was close to the standard arc through the study of the mouth part of bamboo elephant larvae. As a function of the contour curve of the mouth part, a bionic cutting blade was designed. They found that the power consumption of a bionic blade was reduced by 12.8%, and the cutting efficiency was improved by 12.5% when a bionic blade was used to cut cabbage [24]. Therefore, based on the unique morphological structure and geometric characteristics of organisms, the application of bionic elements to the optimal design of agricultural machinery components could achieve better working results.

There are many cutting phenomena in nature; for example, many animals tear, cut or bite plant stalks and leaves. This includes the tea pest cricket: after a long evolution, it has excellent chewing mouth parts, which can effectively cut the stalks of tea trees. Therefore, this paper observed and analyzed the cricket mouthpart, and designed new bionic cutters for tea stalk cutting, which could reduce the cutting resistance and cutting power consumption and improve the picking quality.

2. Curved Structural Characteristics of Cricket Tooth

The cricket mouth part is a chewing mouth part with a compact and firm structure, which could be used for chewing food, digging holes, building nests and hunting enemies. There is one pair of upper jaws in the cricket mouth part [25,26]. The upper jaw has a row of arc-shaped teeth, which is divided into the incisor lobe and molar lobe (Figure 1). The incisor lobe is the arc-shaped part of the tip of the upper jaw, which is mainly used to pierce and tear food. The root structure of the tooth is known as the molar lobe, which is mainly used for chewing and grinding food. The structure of the molar lobe shows a slight protrusion, while the incisor lobe is more prominent. Through natural evolution and selection, the arc-shaped structure of the teeth and the arrangement of the upper jaw of crickets could easily cut and tear plant fibers [25,26]. The curved structural characteristics of the cricket tooth have great importance in reducing drag and consumption, as well as improving the harvest quality of the cutting blade and cutter.

3. Bionic Curve Extraction and Analysis

In order to quantitatively analyze the influence of the structure and shape of different teeth on cutting, the contour curve of the cricket’s upper jaw was fitted and analyzed [27,28]. The curve was divided into five parts: curve a, b, c, d and e, which were used to clearly express the contour curve (Figure 2a).

In Figure 2a, the curves a, c and e have the same upward and downward trend, whereas the curves b and d have no change law. In the upward section, the curve a increases almost in a straight line, while the curves c and e are slightly convex. Furthermore, the peak value of the curve a is much higher than that of other curves. Figure 2b shows the residual value after adjusting the curve. The lower the residual value, the higher the fitting degree. The absolute values of the residual value on all curves are less than 10. Then, the second-order derivative function of five fitting curves is obtained, and the graphic analysis is drawn, as shown in Figure 2c. The second-order derivative value of the curve a continues to decline from 0.1 to −0.3, and then increases to about 0.55. The second-order derivative values of curves c and e slightly increase from the negative numbers, and the second-order derivative values of curves b and d are approximately 0. The second-order derivative function could reflect the concave convex shape of the image. Because the second-order derivative function is greater than 0, the shapes of curves a, c and e are convex outward, which indicates that the contact area between the tooth and the material increases in the cutting process. The teeth corresponding to curves a, c and e might play a key role in the feeding process of the cricket. In Figure 2d, the curvature of the curve a is much greater than that of other curves, showing that the tooth corresponding to curve a is sharp and conducive to cut food.

From the fitted curve image, second-order derivative function and curvature, the tooth corresponding to curve a is the most convex and sharp of the curves a, b, c, d and e, making it the most suitable for cutting food. The tooth corresponding to curve a is the incisor lobe, which could effectively pierce and cut the tea shoot. Therefore, the arc-shaped structure of the incisor lobe confers on it the ability of reducing the cutting resistance. To reduce the cutting power consumption and improve the cutting quality of tea shoot harvest, the structural parameters of the cricket’s incisor lobe are used as bionic elements in the design of the bionic blade and bionic cutter.

4. Bionic Blade Design

The morphology of the incisor lobe of cricket’s upper jaw is illustrated in Figure 3. The quintic polynomial of the profile curve of the incisor lobe is shown as Equation (1). The fitting degree R² of this formula is 0.999, which indicates that it is close enough to the original curve, and there is no need for higher-order polynomials to describe the contour curve of the incisor lobe.

$$\begin{array}{l}y(x)=0.162+2.614x+0.036x^2-0.006x^3\\ -2.61\times {10}^{-6}{x}^4+2.02\times {10}^{-6}{x}^5\end{array}$$

(1)

In order to reduce the machining difficulty of the cutting tool and retain the bionic characteristics, this paper used another method to fit the morphological structure of incisor lobe for comparing and analyzing the fitting effect. The upward and downward sections of the incisor lobe are adjusted by the least square method, as shown in Figure 4. The expression of the upward section is shown as Equation (2), and the fitting degree R² = 0.964. The expression of the downward section is Equation (3), and the fitting degree R² = 0.953. The slopes of the upward and downward section on the approximate line segment are 1.959 and −1.891, respectively. In the Cartesian coordinate system, the corresponding dip angles are 63° and 118°, as shown in Figure 5. The results showed that the linear regression equations of the upward and downward sections could also be used to express the original curve, although their fitting degree was not as high as that of the quintic polynomial.

$$y(x)=2.387+1.959x$$

(2)

$$y(x)=64.129-1.891x$$

(3)

In summary, the quintic polynomial could precisely describe the contour curve of the incisor lobe. When the precision requirement was not high, in order to simplify the machining process, the straight line could also be used to approximate express the contour curve of the incisor lobe.

According to the literature review and actual measurements of the various tea harvesters, some structural parameters of the cutting blade and cutter could be determined [29,30,31]. The tooth height (L), tooth thickness (b), blade edge angle (θ), blade pitch, cutter length and cutter height were 29.0 mm, 1.5 mm, 35°, 45 mm, 488 mm and 45 mm, respectively. For ease of comparative analysis, some structural parameters (L, b and θ) of the bionic blade and the ordinary blade were the same. The structural parameters of the ordinary blade, the bionic blade e and the bionic blade f are presented in Figure 6. The curve of the bionic blade e was developed on the basis of the quintic polynomial of imitation of the structure of a cricket’s incisor lobe. The bionic blade f was the triangular blade that imitated the incisor lobe structure of the cricket. The tooth crest width (a₁) and tooth root width (a₂) of the ordinary blade were 3.5 mm and 13.0 mm, respectively. The tooth root width (n) of the bionic blade e and the tooth root width (m) of the bionic blade f were 27.9 mm and 30.1 mm, respectively. The internal angles (α and β) of the bionic blade f were 63° and 62°. The ordinary blade and the bionic blade were processed as shown in Figure 7. The reciprocating cutters are shown in Figure 8.

5. Cutting Performance Test of Cutting Blade

5.1. Test Methods

The experimental objective was used to analyze the cutting performance by comparing the cutting experiment of the bionic blade and ordinary blade. In this test, the average moisture content of the tea stalk was 71.4%, and the harvest date was April 2020. The harvested tea stalks were new tea shoots, second inter-node, third inter-node and fourth inter-node. A texture tester (Stable Micro Systems, TA-XT2i) was employed to measure and record the cutting performance (cutting force and energy consumed) of the tea stalk. Subsequently, the cutting performance of the blade was analyzed on the basis of the Box-Behnken design principle in the design expert 8.0.6 software [32]. The inter-node number, tea variety and blade type were chosen as three influencing factors. We took three levels for each influencing factor. The experimental indexes were the cutting force and cutting power consumption. The influencing factor code is presented in

Table 1. Coding of factors.

Level	Factor
	Inter-Node Number	Tea Variety	Blade Type
−1	2	Maolv	Ordinary blade
0	3	Zhongcha108	Bionic blade e
1	4	Baicha	Bionic blade f

.

5.2. Test Results and Analysis

The test scheme is a three-factor three-level test, and the test results are shown in

Table 2. Test scheme and results.

No.	Factor			Evaluation Index
	x1	x2	x3	y	w
1	0	1	1	12.025	1.395
2	1	0	0	17.554	3.096
3	0	−1	1	13.626	1.055
4	−1	−1	0	12.161	0.851
5	−1	1	0	13.321	1.777
6	0	0	0	13.626	1.055
7	1	−1	0	11.973	0.714
8	1	1	0	12.937	1.093
9	1	0	0	11.756	0.886
10	−1	0	1	11.286	0.935
11	1	0	−1	12.835	1.344
12	1	0	1	12.869	0.824
13	0	1	0	13.655	1.494
14	−1	0	−1	13.626	1.055
15	0	0	1	13.626	1.055
16	0	0	0	13.626	1.055
17	−1	0	−1	11.961	2.585

. x₁, x₂ and x₃ were factor code values, while y and w were evaluation indexes. The cutting force and cutting power consumption could reflect the cutting performance of the blade. The greater the cutting force and the lower the cutting power consumption, the better the cutting performance of the blade [33,34].

5.2.1. Cutting Force Analysis

The variance analysis of the cutting force is shown in

Table 3. Variance analysis of cutting force.

Source	Sun of Squares	df	Mean Square	F Value	p Value
Model	29.91	9	3.32	15.39	0.0008 **
A-A	9.04	1	9.04	41.84	0.0003 **
B-B	0.41	1	0.41	1.88	0.2128
C-C	3.82	1	3.82	17.68	0.004 **
AB	0.0053	1	0.0053	0.025	0.8796
AC	7.63	1	7.63	35.33	0.0006 **
BC	0.029	1	0.029	0.13	0.7253
A^2	1.07	1	1.07	4.93	0.0618
B^2	7.69	1	7.69	35.61	0.0006 **
C^2	0.38	1	0.38	1.74	0.2285
Residual	1.51	7	0.22
Lack of fit	1.51	3	0.5
Pure error	0	4	0
Cor total	31.43	16
R2	0.9519
Adj. R2	0.89

Note: ** indicates p < 0.01 (extremely significant).

. The F value and p value could be used as indicators of variance analysis. The F value was larger, and the p value was smaller, indicating that the reliability of the analysis results was higher. In Table 3, the F value of the cutting force regression model is 15.39 and the p value is 0.0008 (p < 0.01), indicating that the model is very significant. The determination coefficient R² of the model was 0.9519, indicating that the model could explain 95.19% of the change of response value, and only 4.71% of the total variation could not be explained by this model. The analysis results showed that the regression model could better characterize the relationship between the cutting force and inter-node number, between the cutting force and tea variety, and between the cutting force and blade type.

In the model, the influence of the primary term x₁, x₃, interactive term x_1×3 and secondary term x₂² were very significant (p < 0.01). Other elements had no meaningful effect. The quadratic regression model between the cutting force (y) and the coding value for each influencing factor was obtained as follows.

y = 13.63 + 1.06x₁ + 0.23x₂ − 0.69x₃ + 0.036x₁x₂ − 1.38x₁x₃ − 0.085x₂x₃ + 0.5x₁² − 1.35x₂² − 0.3x₃²

(4)

In the regression equation, the absolute value of the coefficient of each factor represented the ability of this factor to affect the predicted results of the model. In Equation (4), the absolute values of factors x₁, x₂ and x₃ were 1.06, 0.23 and 0.69, respectively. Therefore, the influence of various factors on the cutting force (y) from large to small was x₁, x₃ and x₂, representing the inter-node number, blade type and tea variety, respectively.

5.2.2. Cutting Power Consumption Analysis

The variance analysis of the cutting power consumption is shown in

Table 4. Variance analysis of cutting power consumption.

Source	Sun of Squares	df	Mean Square	F Value	p Value
Model	5.45	9	0.61	3.79	0.0465 *
A-A	0.77	1	0.77	4.83	0.064
B-B	0.0076	1	0.0076	0.047	0.8341
C-C	2.74	1	2.74	17.13	0.0044 **
AB	0.044	1	0.044	0.28	0.6158
AC	0.015	1	0.015	0.092	0.7711
BC	0.000001	1	0.000001	0.0000063	0.9981
A^2	0.88	1	0.88	5.49	0.0516
B^2	0.39	1	0.39	2.41	0.1643
C^2	0.63	1	0.63	3.96	0.087
Residual	1.12	7	0.16
Lack of fit	1.12	3	0.37
Pure error	0	4	0
Cor total	6.57	16
R2	0.8296
Adj. R2	0.6104

Note: ** indicates p < 0.01 (extremely significant), * indicates p < 0.05 (significant).

. The F value of the regression model of the cutting power consumption was 3.79, and the p value was 0.046 (p < 0.05), indicating that the model was significant. The determination coefficient R² of the model was 0.8296, indicating that the model could explain 82.96% of the change of response value, and only 17.04% of the total variation could not be explained by this model. The analysis results showed that the regression model could better characterize the relationship between the cutting power consumption and inter-node number, between the cutting power consumption and tea variety, and between the cutting power consumption and blade type.

In the model, the influence of the primary term x₃ was very significant (p < 0.01). In addition, other elements had no significant impact. The quadratic regression model between the cutting power consumption (w) and the coding value of each influencing factor was obtained as follows.

w = 1.05 + 0.31x₁ − 0.031x₂ − 0.59x₃ − 0.11x₁x₂ − 0.061x₁x₃ – 5 × 10⁻⁴x₂x₃ +
0.46x₁² − 0.3x₂² + 0.39x₃²

(5)

In the regression equation, the absolute value of the coefficient of each factor represented the ability of this factor to affect the predicted results of the model. In Equation (5), the absolute values for factors x₁, x₂ and x₃ were 0.31, 0.031 and 0.59, respectively. Therefore, the influence of various factors on the cutting force (y) from large to small was x₃, x₁ and x₂, representing the blade type, inter-node number and tea variety, respectively.

In conclusion, the tea variety had the least impact on cutting force and cutting power consumption. Although tea variety could affect the physical and mechanical properties such as stalk diameter, node spacing and internal chemical composition, there were few differences in physical and mechanical properties among the different varieties [35,36]. However, in the physical properties, the number of the stalk inter-node had a great influence on its mechanical properties. Generally speaking, with the increased number of stalk inter-node, the cutting resistance and cutting power consumption of the stalk improved. Additionally, the blade type with a bionic blade could reduce cutting resistance and cutting power consumption.

6. Field Cutting Experiment

6.1. Experimental Design

The field cutting experiment was conducted at a Maichun tea farm, Danyang City, Jiangsu Province, in April 2020. The farm was located in the hilly area of the middle and lower reaches of the Yangtze River (latitude 32°01′ N, longitude 119°40′ E). In this farm, the arrangement of the tea tree was north–south, with many dense branches. The vertical spacing of the tea tree was approximately 1.0 m, and the canopy width was about 0.8 m. After the tea shoot was harvested, the average moisture content of the stalk was measured at 73.5%. The harvested variety of the tea sample was Zhongcha 108, and the cutting part was the second inter-node.

When the tea harvester was working, the machine’s forward speed and picking height were hard to control. Therefore, a test bed with adjustable height and speed was designed to cut a single-row of tea shoots (Figure 9). The total length and width of the test bed were 3.0 m and 0.6 m. The test bed included a bench, a DY-IS stepper motor controller, a DM542 stepper motor drive, a stepper motor, and a synchronous belt guide rail slide. The tea harvester was fixed on the test bed, which was moved with the conveyor belt.

The speed ratio of the cutter was the ratio of the forward speed on the tea harvester to the cutting speed of the cutter. According to the literature review, the best speed ratio of the tea harvester was selected at 0.8 [29]. Depending on the adult step size and actual operational requirements, the forward speeds of the harvester were 1.0 m/s, 0.8 m/s and 0.6 m/s, respectively. The corresponding cutting speeds of the cutter were 0.80 m/s, 0.64 m/s and 0.48 m/s, respectively.

The orthogonal test was carried out by selecting the cutter speed ratio (A), cutting angle (B) and cutter type (C) as test factors, and the integrity rate (M) and missing rate (N) of tea shoot as test indexes. In the experiment., each test factor consisted of three level values. The orthogonal table of the three factors with three levels is L9 (3⁴), as shown in

Table 5. Factors and levels of orthogonal experiment.

Level	Factor
	A		B/°	C
	Cutting Speed/m·s−1	Forward Speed/m·s−1
1	0.48	0.60	−3	Ordinary cutter
2	0.64	0.80	0	Bionic cutter e
3	0.80	1.00	3	Bionic cutter f

Note: 0° is parallel to ground plane; 3° is angle 3° between cutter with ground plane, and direction is upward; −3° is angle −3° between cutter with ground plane, and direction is downward.

. The scheme would carry out nine kinds of combination tests to statistically process the results of the integrity rate and missing rate of the tea shoot so as to find the best combination scheme with the level of each influencing factor.

6.2. Test Results and Analysis

Orthogonal test results for field cutting performance on the tea stalk are shown in

Table 6. Orthogonal test results of field cutting performance.

No.	A		B/°	C	M/%	N/%
	Cutting Speed/m·s−1	Forward Speed/m·s−1
1	0.48	0.60	−3	Ordinary cutter	90.94	5.18
2	0.48	0.60	0	Bionic cutter e	90.53	5.13
3	0.48	0.60	3	Bionic cutter f	81.60	5.61
4	0.64	0.80	−3	Bionic cutter e	93.13	3.57
5	0.64	0.80	0	Bionic cutter f	96.04	5.04
6	0.64	0.80	3	Ordinary cutter	80.50	7.07
7	0.80	1.00	−3	Bionic cutter f	94.41	3.99
8	0.80	1.00	0	Ordinary cutter	93.01	4.15
9	0.80	1.00	3	Bionic cutter e	87.64	4.58
Integrity rate	Index	A	B	C
	Mean value-1	87.689	92.827	87.816
	Mean value-2	89.558	92.524	90.433
	Mean value-3	91.351	83.247	90.348
	Range	3.662	9.580	2.284
	Factor order	B > A > C
	Better combination	A3B1C2
Missing rate	Index	5.307	4.247	5.466
	Mean value-1	5.227	4.771	4.428
	Mean value-2	4.238	5.755	4.879
	Mean value-3	1.069	1.508	1.039
	Range	5.307	4.247	5.466
	Factor order	B > A > C
	Better combination	A3B1C2

Note: values of 1, 2 and 3 in better combination are respectively represented as levels of 1, 2 and 3 in Table 5.

. Intuitive analysis and variance analysis were performed on the orthogonal test results to investigate the influence of the test factors on the test indicators, and a better combination of factors was achieved.

The intuitive analysis results of the integrity rate and missing rate of the tea shoot are presented in Table 6. The range analysis could determine the influence order of the experimental factors on the integrity rate and missed rate of the tea shoot. The greater the integrity rate of the tea shoot, the better the results [33]. The lower the missing rate of the tea shoot, the better. The influence order on the integrity rate and missing rate of the tea shoot from large to small was B > A > C, representing the cutting angle, cutter speed ratio and cutter type, respectively.

To determine the influence degree of the test factors on the index, the results of the orthogonal test were analyzed by the variance (

Table 7. Variance analysis for integrity rate and missing rate of tea shoot.

Integrity Rate				Missing Rate
Factor	Freedom	Sum of Squares of Deviations	F Value	Factor	Freedom	Sum of Squares of Deviations	F Value
A	2	20.11	1.56	A	2	2.13	1.57
B	2	177.94	13.79 *	B	2	3.52	2.61
C	2	13.27	1.03	C	2	1.63	1.21
Error	2	12.91		Error	2	1.35
Total	8	224.23		Total	8	8.62

Note: * indicates significance at the level of 0.1.

). In the variance analysis, the F value indicated the influence significance of the test factors on the index. In Table 7, the significant order for the integrity rate of the tea shoot is the cutting angle, the cutter speed ratio, and the cutter type. The significant order for the missing rate of the tea shoot was the cutting angle, the cutter speed ratio, and the cutter type. The significant order for the integrity rate and missing rate of the tea shoot was the same, which was similar to the intuitive analysis conclusion.

In Table 7, it can be observed that the integrity rate of the tea shoot increases with increases in the cutter speed ratio, decreases with increases in the cutting angle, and changes with changes in the cutter type. The missing rate of the tea shoot decreased with an increase in the cutter speed ratio, increased with an increase in the cutting angle, and changed with a change in the cutter type. Due to the tea stalk having a high toughness and a certain angle of inclination, the appropriate cutting angle would cause the cutter to slide on the stalk. When the cutting angle was 3°, the cutter would slide over the stalk skin. When the cutting angle was 0°, the cutting loss might also be caused by the inclination of the stalk. Thus, the cutting effect at a cutting angle of −3° was better than that of the cutting angle of 3° and 0°. Moreover, the bionic cutter had great advantages in reducing cutting resistance and cutting power consumption, as well as in improving cutting quality. According to the intuitive analysis and variance analysis, the cutting angle was the most important to the integrity rate and missing rate of the tea shoot, followed by the cutter speed ratio, and finally the cutter type. In this study, the optimal parameter combination within the range of test factors was A₃B₁C₂, that is, the cutting speed of 0.8 m/s, forward speed of 1.0 m/s, cutting angle of −3°, and bionic cutter e.

6.3. Parameter Combination Optimization and Verification

In order to verify the effect of the bionic cutter, a comparative test under the condition of the optimal parameter combination and other combinations was carried out to cut a single row of tea shoots. When the bionic cutter and −3° cutting angle were not used, the integrity rate and missing rate of the tea shoot were 79.5% and 10.3%. In the condition of optimal parameter combination (cutting speed of 0.8 m/s, forward speed of 1.0 m/s, cutting angle of −3°, and bionic cutter e), the integrity rate and missing rate of the tea shoot were 92.7% and 3.9%. This showed that measures to optimize the combination of parameters could improve the quality of the harvest (increased by integrity rate 13.2% and decreased by missing rate 6.4%).

7. Conclusions and Discussion

In the process of crop harvesting, the harvest quality and energy consumption are worthy of attention [37,38]. With the improvement in mechanized harvesting level, these problems have been paid more and more attention in recent years. Due to the complexity of the natural environment and the particularity of crops, the harvest of different crops can vary greatly [39]. As already mentioned, for different crops, many scholars have put forward the methods of saving energy, reducing consumption and improving harvest quality to develop new harvesting systems [40,41], such as the use of bionic principles to design bionic cutters. Therefore, a new bionic blade and bionic cutter for a single-row tea harvester has been developed on the basis of the structure and motion parameters of the cutter, which could improve the quality of the harvest.

In this study, the bionic structure optimization of a reciprocating cutting system was carried out based on the arc-shaped structure of the incisor lobe of a cricket. The cutting performance test showed that the bionic blade could effectively reduce the cutting resistance and cutting power consumption. Moreover, the optimum combination of a 0.8 m/s cutting speed, 1.0 m/s forward speed, −3° cutting angle and bionic cutter e was determined with a field cutting experiment. In the optimal parameter combination, the integrity rate and missing rate of the tea shoot were 92.7% and 3.9%, respectively. For this reason, the method of a bionic blade and bionic cutter on tea proposed in this paper could provide technical reference for high-quality cutting of a single-row tea harvester.