Pei Fengque, Yang Kaiwei, Wang Miao, Tong Yifei*

(1. College of Mechanical and Electrical Engineering, Hohai University, Changzhou 213022, China；2. School of Mechanical Engineering, Nanjing University of Science & Technology, Nanjing 210094, China)

Abstract: Intelligent and green agricultural machinery has become an important goal at this stage.The application of intelligent manufacturing in the field of agricultural machinery has ushered in unprecedented market opportunities.This paper firstly analyzes the development status of agricultural machinery, and studies the key enabling technologies for intelligent manufacturing in agricultural machinery, including new IT, big data, cloud computing, digital twin and robotics, etc.Then the new development and application brought by intelligent manufacturing to agricultural machinery are summarized, including the design, management and maintenance, operation and new energy of agricultural machinery.Finally, the challenges and application prospects are discussed, and urgent problems are pointed out, including the lack of basic data, limitation of intelligent manufacturing in agricultural machinery, insufficient intelligent operation and maintenance as well as less application of intelligent operation in agricultural machinery.It is suggested that in the future, four research perspectives should be noticed, namely the design of knowledge base, the breakthrough of key application, the establishment of a unified platform and the intelligence of agricultural machinery.By solving these bottlenecks, the development of intelligent manufacturing in agriculture machinery will be greatly promoted.

Keywords: intelligent manufacturing；agricultural machinery；new IT；big data；design and manufacture；operation and maintenance

0 Introduction

As the most basic industry in national economy, agriculture is not only closely related to people’s food, clothing and living standards, but also the basis for the development of other industries.China has a large population and limited resources.Improving agricultural productivity has always been the strategic focus in the process of modernization.Currently China accounts for less than 10% of the world’s cultivated land, which has been producing 25% of food and feeding 20% of the population in the world at large[1].Agricultural machinery, as an important auxiliary equipment for production and an important means to improve productivity, has made a great contribution in this field.

The national agricultural mechanization development plan for the 14th Five-Year plan issued by the Ministry of Agriculture and Rural Affairs in January 2022 has mentioned that aiming at the needs of agricultural mechanization, it is necessary to accelerate the innovation of agricultural machinery and equipment, promote the technical integration of agricultural machinery systems and realize the information management and scheduling of important agricultural mechanized production.As a result, intelligent and green agricultural machinery acceleration has become an important goal.

Under the background of “intelligent and green agricultural machinery” in China, this paper analyzes the development status of agricultural machinery, studies the key integration technologies of intelligent manufacturing and agricultural machinery, and discusses the challenges and application prospects of intelligent manufacturing in the field of agricultural machinery.The second part of this paper analyzes the development status of agricultural machinery.The third part specifically studies the key enabling technologies of intelligent manufacturing.The fourth part summarizes the application status of intelligent manufacturing in the field of agricultural machinery, and last one discusses the challenges and application prospects.

1 Development of agricultural machinery

The growth of Chinese agricultural machinery in 2020 is shown in Table 1.In 2020, the total power of Chinese agricultural machinery was 1.056 billion kW, an increase of 2.79% over 2019 and 17.07% over 2015.All kinds of agricultural machinery increased steadily, of which the largest increase was unmanned aerial vehicle for plant protection(more than 77%).The development of agricultural machinery mainly presents the following characteristics.

Table 1 Summary of China’s agricultural machinery ownership in 2020

(1)The total amount of agricultural machinery continued to grow steadily.By 2020, the comprehensive mechanization rate of crop cultivation and harvest in China had reached 71.25%, with an increase of 1.23% over 2019.However, there is still a gap compared with the world advanced level.It is necessary to continue to increase the total amount of agricultural machinery and promote agricultural mechanization in China.

(2)The trend of transformation and upgrading of agricultural machinery to high quality and efficiency is obvious.As agricultural production enters a new stage of mechanization, the demand structure of agricultural mechanization in various fields of agricultural production has undergone profound changes, and the contradiction of unbalanced and insufficient development of agricultural mechanization in regions, industries, varieties and links has become prominent, which puts forward higher requirements for agricultural machinery.It is necessary to promote the all-round development of agricultural mechanization and transformation to high quality and efficiency.

The intellectualization and greening of agricultural machinery is a kind of difficult and complex system engineering, and intelligent manufacturing provides a solid technical support for its realization.intelligent manufacturing is a new production mode based on the deep integration of novel information and communication technology and advanced manufacturing technology, which runs through all links of manufacturing activities such as design, production, operation and maintenance, service and so on.It has the functions of self-perception, self-learning, self-decision-making, self-execution, self-adaptation and so on[2].Specifically, intelligent manufacturing includes two meanings.The first one is “intelligence”, which includes novel information and communication technologies and advanced manufacturing technologies such as artificial intelligence(AI), machine learning(ML), digital twins(DT), cloud computing, big data, Internet of things and 5G[3-7]; The second one is manufacturing, including manufacturing R & D, design, supply chain, production, operation and maintenance, service and other links, as well as various production factors of human, machine, material, method, environment and manufacturing process, and green manufacturing based on industrial chain collaboration[8-9].As for the practical aspect, intelligent manufacturing refers to the application of various new technologies in the whole value chain and production process of manufacturing industry.The state and government have promoted the transformation and upgrading of agricultural machinery to intelligent and green manufacturing by administrative guidance.Intelligent manufacturing has witnessed a vast market opportunity in the field of agricultural machinery[10-11].

2 Key enabling technologies of intelligent manufacturing

2.1 Agricultural decision technology driven by new IT

New IT is a new generation information technology, including ML, AI, blockchain and other technologies, enabling the transformation and upgrading of all walks of life.In the field of agricultural machinery, new IT plays a vital role in decision-making technology.Three core technologies represented by ML, deep learning and various heuristic algorithms transform the traditional decision-making means that strongly rely on human experience into a more scientific AI decision-making method[12].

With the new technology era of AI, ML has been proved to be a powerful tool for data analysis, and its application in the field of agricultural machinery is also in increasing demand.Based on PCA dimension reduction and SVM, Tang[13]analyzed the principle of road condition recognition of agricultural vehicle auto drive system, and designed the hardware platform of auto-drive system with modular technology.Finally, he realized the agricultural vehicle auto-drive system based on embedded learning and ML.

Deep learning is a new research direction in the field of ML.It has made many achievements in image processing, speech recognition and so on.Many scholars also apply it to the field of agricultural machinery.Zhang et al.[14]proposed an AMT net network to re-recognize the image of agricultural machinery.The experimental results showed that this network was better than the traditional ResNet 50 and Inception-v3 network, and features better performance and good robustness to illumination, environmental change and small area occlusion, which met the actual requirements of intelligent monitoring of agricultural machinery operation.Aiming at the problems of high failure rate and low operation efficiency of combine harvester, Xi[15]proposed a fault diagnosis system of combine harvester based on deep learning.This method combined the improved stacking self-encoder and multi-sensor information fusion technology, which could realize the effective detection, intelligent diagnosis and accurate evaluation of the operation state of combine harvester.Inspired by the idea of end-to-end control in the field of automatic driving, He[16]studied the visual navigation and control of intelligent weeding robot.It combined CNN end-to-end network and RNN network structure to control the intelligent weeding robot driving along the crop line.

Heuristic algorithm is defined in opposite to optimization algorithm.It is an empirical exploratory algorithm, mainly including genetic algorithm, ant colony algorithm, simulated annealing algorithm and so on.In recent years, many scholars have applied it to the path planning and scheduling of agricultural machinery, and solved the corresponding problems.Sethanan et al.[17]believes that the job path planning of the harvester has a great impact on improving the harvesting efficiency of sugarcane machinery.In his research, he creatively designed a particle coding and decoding scheme, which combined the accessibility of the plot and constraints for segmented harvesting with path planning, in order to solve the path planning problem of sugarcane harvester.Duan et al.[18]proposed an adaptive fuzzy control algorithm based on genetic algorithm, to solve the problem of establishing an accurate mathematical model for the navigation system of agricultural machinery.Through the data fusion method, the lateral deviation and heading deviation sensors were fused into one data to control the parameter, and finally made the agricultural machinery move along the desired path.Cerdeia-Pena et al.[19]took forage harvesting as the research object, and uses simulated annealing algorithm and tabu search algorithm to solve the problem of agricultural machinery allocation.Seyyedhasani et al.[20]also used tabu search meta heuristic algorithm compared with CW algorithm, and simulated the real working scene of agricultural machinery by changing the number and path of agricultural machinery working.

The experimental verification showed that tabu search algorithm was more efficient than CW algorithm when allocating multiple agricultural machinery at the same time.

New IT technology represented by ML, deep learning and various heuristic algorithms plays an important role in agricultural machinery decision-making system, and provides a theoretical basis and feasible model for the navigation system, fault diagnosis, intelligent monitoring, and operation time of agricultural machinery.The specific application technology is summarized in Table 2.

Table 2 Application of new IT technology in agricultural machinery

2.2 Data driven decision technology

Due to the enormous amount of data generated in the process of agricultural machinery design or operation and maintenance, the concept of big data and other analysis methods need to be introduced.In recent years, with the development of IoT, computer technology, AI and other technologies, these data can be analyzed and found their value, which promotes the development of decision-making technology in the field of agricultural machinery and improves the performance and application effect of agricultural machinery.

Data driven smart manufacturing technology is widely used in agricultural machinery design.For example, it took the IoT technology into the design of agricultural machinery assembly and manufacturing system[21], and adopted neural network analysis algorithm to speed up the data analysis, which improved the query ability of agricultural product parts manufacturing information, and effectively shortened the design cycle of agricultural machinery gearbox and reduced the manufacturing cost.In order to improve the accuracy of feature extraction of agricultural machinery CAD model, Liu et al.[22]proposed a CAD model feature extraction and evaluation method based on three-dimensional wavelet transform, which could shorten the design cycle, meet the needs of designers and promote the intelligent design level of agricultural machinery.Based on the traditional design mode, Zhang et al.[23]aims at the problems of complex design methods and difficult time control of agricultural machinery products, proposed an agricultural machinery digital cloud platform based on big data collaborative computing processing and distributed storage principle technology, which could effectively improve the model optimization design efficiency and the process optimization design efficiency of agricultural machinery parts.

In the operation and maintenance of agricultural machinery, data driven technology has also attracted much attention.For example, Wang et al.[24]introduced machine vision technology into the image recognition system of apple sorter, and realized the grade classification of apple and the automation of apple sorting through the data collection, processing, feature extraction and calculation of apple pictures.Aiming at the problem of agricultural machinery fault analysis, Li et al.[25]established a new agricultural engineering machinery management and analysis platform by collecting and analyzing the ground data of agricultural engineering machinery, with the help of IoT and big data technology, which has been applied in many farms in China.

The data-driven decision-making technology, which mainly includes feature extraction and big data analysis, has provided a strong guarantee and basis for the decision-making in design or operation and maintenance of agricultural machinery.At the same time, it effectively improves the design efficiency of agricultural machinery products, reduces the manufacturing cost of products and increases the accuracy of operation and maintenance.

2.3 Cloud based computing and storage strategy

With the rapid development of multi-source heterogeneous acquisition network technology, the growing number of agricultural machinery intelligent devices is producing a large amount of data that need to be calculated, stored or processed in real time, which is difficult to achieve only by a single server.Cloud technology with cloud platform, cloud computing and cloud edge or fog technology as the core, integrates a variety of computer, AI and virtual simulation technologies, providing technical support for cloud computing and cloud storage of agricultural machinery data.Cloud computing is also known as network computing, where “cloud” means that the services provided are on the Internet.Its schematic diagram is shown in Figure 1.

Figure 1 Cloud computing diagram

In recent years, many scholars have applied cloud computing, cloud platform and other technologies to the field of agricultural machinery, which has improved the intelligent level of agricultural machinery production, manufacturing and operation.To improve the efficiency and accuracy of the picking robot positioning system, took the cloud platform technology and computer control system into the design of the positioning system to improve the accuracy of image feature extraction and positioning of the picking robot[26].In order to improve the intelligent level of design and manufacturing of agricultural machinery parts, Zhang et al.[27]introduced the smart manufacturing system into the design and production process of agricultural machinery parts based on cloud platform and virtual simulation software, which improved the virtual simulation of parts production process.Zhang et al.[28]and others proposed a rapid design method for modular agricultural machinery products based on cloud computing and three-dimensional digital design concept to solve the problems of product diversity, personalization and short order delivery cycle in modern agricultural machinery market.

With the help of cloud computing technology to build a private cloud of agricultural machinery and link agricultural machinery together through IoT technology, it is easier to realize the cross regional operation intelligent scheduling function of agricultural machinery, including agricultural machinery location query, tracking and viewing, information receiving and publishing, agricultural machinery condition monitoring, agricultural machinery fault remote diagnosis, etc., which greatly improves the application effect and guarantee of agricultural machinery.

2.4 Virtual real mapping technology driven by digital twins

Digital twin technology(DT)has become a hot spot in the current academic and industrial circles.Figure 2 is a schematic diagram of DT.Based on sensor technology, internet of things technology, virtual reality technology and other multidisciplinary technologies, it establishes a digital model of a physical entity and completes the mapping in the virtual space, so as to reflect the whole life cycle process of the corresponding physical equipment.The application of digital twin technology to the design, production, operation and maintenance of agricultural machinery can effectively improve the production efficiency and the intelligence of equipment fault diagnosis.

Figure 2 Digital twin diagram

Akamker et al.[29]proposed a method of DT in potato picking, so that the speed of potato harvester and conveyor belt could cooperate with each other, reduce the damage caused by potato falling, and improve the quality of agricultural products.Nemtinov et al.[30]proposed a method to create a DT model of complex agricultural machinery, and gave its framework and corresponding examples: a combined unit for soil preparation and cooling of grain crops and a unit for cleaning and calibration of grain seeds.

The deep integration of DT and agricultural machinery will provide new functions for the digital transformation and upgrading of agricultural machinery, which plays an important role in intelligent cultivation and intelligent control of agricultural machinery, and promote the intelligent development of agricultural machinery in China.

2.5 Robotics

Due to the rough working environment of agricultural machinery, and the high repeatability of operation, human error and quality fluctuation are inevitable.While agricultural robot is an advanced form of agricultural machinery, integrates all kinds of intelligent equipment and advanced technology, and can work instead of technicians under rough conditions.

Zhao et al.[31]developed an agricultural intelligent spraying robot based on big data.They integrated the collected robot operation information into big data and then analyzed it.The robot can realize autonomous navigation and accurate spraying, and its operation effect and intelligent level is improved.Mao et al.[32]designed a small orchard transportation robot with dual navigation mode to solve the problems of single autonomous navigation mode of transportation equipment after orchard apple picking and unable to start or stop at any point.It can choose pedestrian guidance navigation or fixed-point navigation according to the demand, which meets the functions of autonomous transportation and safe obstacle avoidance of orchard robot.Based on an adaptive neuro fuzzy information system, Tian et al.[33]developed a sensitive sliding sensor with a piezo resistor to accurately control the griping force of the agriculture robot, which can prevent fruits or vegetables from mechanical damage caused by agriculture robots.Agricultural robots can replace technicians to operate automatically under various working conditions, which effectively improve the automation and intelligence level of agriculture.

3 Smart manufacturing enables new fields of agricultural machinery

3.1 Agricultural machinery product design based on Intelligent Technology

The traditional design method of agricultural machinery has low efficiency, heavy workload, is easy to make mistakes, and the design effect is not intuitive.Moreover, the optimization of parts after trial production of prototype seriously prolongs the design cycle.The design process of agricultural machinery products adopts intelligent design method, combined with the actual working conditions of agricultural machinery and modern Internet and intelligent technology in traditional design, which is conducive to the accuracy and efficiency of the design process of agricultural machinery products[34].

Guan et al.[35]introduced the scheme of self-learning resource sharing and exchange interaction under the multimedia network environment into the construction of agricultural machinery digital design platform.By digital design, the virtual sample machine of parts can be obtained without actual processing via the simulation of processing process.Through the simulation test of the performance of virtual prototype, the optimization of parts can be realized.It took the network into the modeling design process of agricultural machinery[36], and output the operation animation of agricultural machinery through the modeling and animation simulation, so as to provide basis for the optimal selection of agricultural machinery modeling.To establish the model of parts, used the interactive design principle to carry out virtual assembly of parts, and obtained the overall model of gearbox, which realized the remote intelligent collaborative design and concurrent design of agricultural machinery, effectively shortened the design cycle of parts.The combination of intelligent technology and agricultural machinery product design improves the collaborative ability of designers and greatly shortens the development cycle of agricultural machinery[37].

In the design of agricultural machinery products, VR technology also shows good application effects.Based on CAD, the most common application is to enhance the content of the visualization system, independently create a functional model, and transfer it to the VR environment with the help of model input form to achieve the goal of enhancing products.There is a special VR helmet display instrument in the design of agricultural machinery products, which allows the operator to directly generate the control experience through the screen.Therefore, the internal function of VR-CAD can be fully integrated into the virtual design environment of large-scale products.

3.2 Operation and maintenance of agricultural machinery products based on intelligent technology

In the process of traditional agricultural production, agricultural machinery is facing a complex agricultural production environment, which is prone to various sudden failure problems.Moreover, with the increasingly complex structure of agricultural machinery, the possible failure problems of agricultural machinery cannot be captured in time, resulting in the damage and delay, which increases the cost of maintenance.

The operation and maintenance of agricultural machinery involves information acquisition, efficient scheduling, remote operation and maintenance, fault early warning, intelligent diagnosis and others, which strongly depends on various enabling technologies of smart manufacturing, such as sensor technology.When the sensors in the agricultural machinery system detect abnormalities such as pressure, temperature and speed in the working process of agricultural machinery, or through the comprehensive analysis of data obtained by various sensors, it can be concluded that there are faults in some important work parts.The early warning function components are controlled by data transmission to perform corresponding early warning functions.For advanced agricultural machinery equipment, when the system determines that there is a fault problem with the machines and tools, it may also implement active shutdown or local shutdown through the control system to protect the important parts and tools.

In addition, CAN technology has been widely used in agricultural machinery, but there is a lack of remote monitoring system combining CAN technology with IoT.Wen et al.[38]studied a remote monitoring system of corn cultivator variable rate fertilizer applicator based on open platform OneNet.The ST32103 main controller is used for data processing and conversion, and the BC20 wireless communication module is responsible for data transmission.With the help of OneNet platform, the PC end and mobile end can realize the real-time remote monitoring of the speed, coordinates, fertilizer discharge shaft speed and other state parameters of the fertilizer applicator.

In addition, in order to solve the problem of high power consumption of wireless communication sensor nodes of agricultural machinery under harsh conditions and improve the overall performance of engine monitoring system, such as solar engine monitoring system based on information acquisition by intelligent agricultural and IoT.The system can measure the voltage, current and temperature of photovoltaic cells and the speed, temperature and noise of the engine, and make early warning in time according to the measured value to realize remote control[39].

3.3 Picking based on intelligence and vision

China is a big country in the production and export of agricultural products.In recent years, the planting area of agricultural products in China has become larger and larger, but most of the production links still rely on labor, especially in the harvest link.At present, there is a sharp shortage of agricultural labor force, which limits the development of the industry to a great extent[40].The actual demand promotes the development of picking technology based on intelligence and vision.In view of the problem that the impurity content or crushing rate is too high during the operation of the combine harvester, Chen et al.[41]proposed an on-line identification method of rice broken grains and impurities in the combine harvester.Using the scheme of collecting rice images in the flow, an image acquisition device was developed to collect rice images in the flow state in real time, and then use OpenCV for image processing to identify and classify according to color and area characteristics of complete grains, broken grains and impurities in rice.After testing, this method can identify the impurities online such as complete grain, broken grain, rice stalk and stem, and provide technical support for the online automatic regulation of operation parameters of rice combine harvester.In view of the phenomenon that the cutter of sugarcane harvester cannot automatically adjust according to the change of sugarcane terrain, Ma et al.[42]designed and established an automatic control system for sugarcane harvester cutting depth that indirectly reflects the cutting depth of sugarcane harvester with the load pressure signal of the system.The system can be adjusted to the cutting depth of about 20 mm with the change of load pressure.The applicability and reliability of the system are verified through the garden test operation.The adjustment error is about 2 mm, which is in line with the results of systematic error analysis.

3.4 Navigation and driverless technology of agricultural machinery

Autonomous navigation technology is the most important technology in driverless technology.Through the computer system inside the agricultural vehicle to control the vehicle and calculation information, and the on-board sensor to obtain real-time environmental information and the high-precision positioning system to locate the vehicle in real time, the agricultural machinery can realize safe auto-driving in an unknown environment without human operation.

Aiming at the problem that the tractor may encounter positioning failure or excessive deviation due to abnormal fluctuation or loss of sensor information in the complex field working environment, Zhu et al.[43]designed a multi-sensor fusion positioning system based on confidence weighting.The system is mainly composed of GPS positioning system and dead reckoning system.The confidence of each sub filter is calculated based on the confidence distance and confidence function of each sensor.After the weighting operation, it is used as an adaptive factor for optimal fusion in the full filter to obtain more accurate pose estimation.

In order to solve the problems of poor human-computer interaction experience and low degree of informatization and intelligence of the existing agricultural machinery navigation system in China, such as a set of navigation management systems based on Android[44], including agricultural machinery operation parameter management, farmland geographic information management, operation path planning Navigation real-time monitoring and historical job data management and other functional modules.The system can achieve stable and reliable operation, good human-computer interaction experience, and can effectively realize the management and monitoring functions in the automatic navigation of agricultural machinery.Agricultural machinery auto drive system is one of the key technologies of unmanned farm, and also an important technical support for cost reduction and efficiency increase of precision agriculture.

Duan[45]built the hardware system of unmanned and intelligent operation based on the traditional rice direct seeding machine, studied and realized the scheme of high-precision positioning and navigation system suitable for paddy field operation environment, automatically upgraded the steering system, speed control system and sowing operation system of the traditional rice direct seeding machine, and realized the long-distance wireless communication scheme.Guo[46]developed the automatic steering wheel of the transplanter, refitted the transplanter, designed the data processing controller and steering controller, selected John Deere high-precision GPS receiver, constructed the driverless control hardware system, and verified that the hardware system can realize the driverless control of the transplanter.

In 2020, Tianjin High End Equipment Research Institute of Tsinghua University and National Agricultural Machinery Equipment Innovation Center jointly developed the first 5G + hydrogen fuel driverless tractor in China.As reported about the super tractor No.1, it can use 5G communication technology for remote control, real-time sensing the running state of the tractor and the surrounding working environment, and carry out accurate work.

According to review, the navigation and driverless technology has realized the effective integration and control of agricultural machinery automatic navigation operation system.While the big ten key technologies are remain to be consumed, such as the navigation and positioning, path tracking, electro-hydraulic steering, motor steering, speed by wire control, machine control, automatic obstacle avoidance, master-slave navigation, on-board terminal and system integration.After the high-precision continuous and stable positioning and attitude measurement of agricultural machinery under different operating conditions, the tracking accuracy of agricultural machinery must be improved greatly.

3.5 New energy technology

Agricultural machinery plays an important role in agricultural production, but the use of a large number of diesel engines consumes a lot of fuel resources and pollutes the environment.With the development of new energy technology, agricultural machinery products powered by new energy have been focused on research and development in recent years, making new energy technology more widely used in agricultural machinery, such as “super tractor 1”, BYD “future agricultural cooperatives, family farms, manors”.These are good examples to drive agricultural machinery with new energy technology, aiming at hilly landform, various greenhouses, orchards, and nurseries.The operation environment in narrow areas such as tea garden can replace the traditional diesel and gasoline engine as the power output to reduce the operation cost.New energy agricultural machinery overcomes the disadvantages of traditional agricultural machinery, such as easy damage, high maintenance cost and short service life.It has lighter body, better performance and higher precision.It is an important battlefield for the development of agricultural machinery in the future.

4 Challenges and prospects of smart manufacturing in the field of agricultural machinery

Although some achievements have been made in the application of smart manufacturing in the field of agricultural machinery, due to the late start of research in this field in China, there are still a series of problems to be solved.

(1)The intelligent design of agricultural machinery lacks basic data.The basic data of agricultural machinery is a necessary condition for the design of agricultural machinery based on smart manufacturing.However, due to the differences in China’s agriculture in regions, industries, varieties and links, some industrial varieties, cultivation and equipment are not matched, the planting and breeding methods, post-planting processing and mechanized production are not coordinated, and the development of relevant agricultural machinery is insufficient, which has a great impact on the promotion of intelligent design of agricultural machinery.

(2)The smart manufacturing of agricultural machinery is not profound enough.It is mainly reflected in two aspects: Firstly, at present, smart manufacturing technology is only applied in some agricultural machinery manufacturing processes, and has not been popularized and standardized in the whole field of agricultural machinery; Secondly, only some smart manufacturing technologies have been applied, and the degree of application is not profound.Some smart manufacturing technologies have not found the entry point of application in agricultural machinery manufacturing.

(3)There are still deficiencies in the intelligent operation and maintenance of agricultural machinery.At present, the application of agricultural machinery operation and maintenance technology based on smart manufacturing is relatively mature, but there are still deficiencies.It is mainly reflected in two aspects.One is that the state data acquisition technology used to support the operation and maintenance of agricultural machinery still needs to be strengthened.At present, there are still some data that lack effective acquisition means, and the data that can be collected are also vulnerable to interference, resulting in uncertainty; Secondly the operation and maintenance lacks a unified platform for system integration, which is unfavorable to promoting the serialization and industrialization of intelligent operation and maintenance of agricultural machinery.

(4)The application of intelligent operation of agricultural machinery is not enough.At present, the intelligent operation of agricultural machinery based on smart manufacturing mainly focuses on planting, irrigation, plant protection and picking, and has carried out relevant research on its key technologies such as path planning, automatic obstacle avoidance, unmanned driving and remote control.However, most of the existing research is in the experimental stage, and there are few practical applications of large-scale planning.

With the development of IoT, computer technology, AI and others, the process of smart manufacturing has been continuously promoted, and the influence of smart manufacturing technology on the field of agricultural machinery will continue to increase, ushering in a new round of development as follows.

(1)Build a design knowledge database to promote the intellectualization of agricultural machinery design.In view of the lack of design data caused by the current incomplete and unbalanced development of agricultural machinery, it is necessary to promote relevant research, build a design knowledge database in the development process of agricultural machinery in different regions, industries, varieties and links, and promote the intelligent process of agricultural machinery design.

(2)Seek the breakthrough of application and promote the intellectualization of agricultural machinery manufacturing.On the one hand, find the bottleneck and solve it through smart manufacturing related technologies to effectively improve production quality and efficiency; On the other hand, the smart manufacturing technology used in the manufacturing process of agricultural machinery should be studied to deepen the application degree of technology and improve the application efficiency.

(3)Establish a unified platform to promote the intellectualization of agricultural machinery operation and maintenance.Carry out system integration on the agricultural machinery operation and maintenance technology currently in use, provide users with a unified platform and improve use efficiency.At the same time, further the study of data acquisition technology, improve the efficiency and quality of data acquisition, and provide data guarantee for agricultural machinery operation and maintenance platform.

(4)Promote the practical application of intelligent operation of agricultural machinery.Strengthen the research on the key technologies of intelligent operation of agricultural machinery, solve the current bottleneck problems, and promote the process of intelligent operation from the experimental stage to the production and application stage, so as to make practical contributions to China’s agricultural development.

5 Conclusions

The new generation information and communication technology and advanced manufacturing technology contained in smart manufacturing play a vital role in promoting the design, manufacturing, operation and maintenance of agricultural machinery.At the same time, smart manufacturing is also an important means for the all-round development of agricultural machinery, transformation and upgrading to high quality and efficiency, and improving the level of intelligence and greening.

Firstly, this paper analyzes the current situation of the development of agricultural machinery in China, and defines the needs of the intelligent development of agricultural machinery.Secondly, it studies the key enabling technologies of smart manufacturing, including new IT, big data, cloud computing, DT and robot technology.The application and bottlenecks of each technology in agricultural machinery are analyzed.Then it summarizes what new changes smart manufacturing has brought to the field of agricultural machinery, including the design, operation and maintenance, picking, navigation, auto-driving and new energy.Finally, the challenges and application prospects of smart manufacturing in the field of agricultural machinery are discussed.It is pointed out that smart manufacturing in the field of agricultural machinery still needs to solve the problems such as the lack of basic data in the intelligent design of agricultural machinery, the lack of in-depth smart manufacturing of agricultural machinery, the lack of intelligent operation and maintenance of agricultural machinery and the less application of intelligent operation of agricultural machinery.

Although there are still some problems in the application of smart manufacturing in the field of agricultural machinery, the current powerful role of smart manufacturing in promoting the intellectualization and greening of agricultural machinery shows the potential of smart manufacturing to the society.It is believed that in the near future, the field of agricultural machinery will usher in great development under the promotion of smart manufacturing.