Research progress on the monitoring methods of the separating loss in grain combine harvester*

Yang Li, Zhen Xue, Lizhang Xu, Yaoming Li, Jie Qiu, Yingfeng Wang

(1. College of Agricultural Equipment Engineering, Jiangsu University, Zhenjiang, 212013, China;2. Taizhou Xiechuang Agricultural Equipment Co., Ltd., Taizhou, 225300, China)

Abstract： The grain loss during the operation of combine harvester is directly related to the economic benefits of farmers. Mastering the accurate real-time grain loss information not only provide intuitive basis for the manipulator to adjust the working parameters of combine harvester that ensures the operation quality and improves the operation efficiency but also facilitate farmers or relevant government departments to track, supervise and measure the harvest loss of each region. The grain loss mainly includes the header loss, the unthreshed grain, the separating loss, the cleaning loss and leakage loss. The cleaning loss and separating loss account for a large proportion of grain loss. Therefore, it is necessary to monitor them in real time. Chinese researches on the monitoring methods of the separating loss are still at the theoretical and experimental stage, which is different from the researches oversea. Methods to improve the accuracy, reliability and adaptability of the separating loss sensors are difficult and essential in the information developing process of the grain combine harvester in our country. This paper summarizes the research progress of the monitoring methods and devices of grain combine harvester’s separating loss at home and abroad from the aspects of the monitoring principles of the separating loss, the sensor signal processing system, the sensor damping method, the probability distribution model of separating loss. In addition, the developing trend will be analyzed. This will contribute to the continuous improving and upgrading of separation loss sensor and lay a solid foundation for the completion of information harvest in China .

Keywords: combine harvester; separating loss monitoring; research progress

0 Introduction

Grain combine harvester is mainly used for harvesting rice,wheat, corn, millet and other food crops as well as some economic crops such as soybean and rape. It has multiple functions such as cutting, threshing, separation, cleaning, bagging or unloading grain in one operation[1]. Grain loss rate is one of the most important indicators when measuring the performance of grain combine harvester, which is directly related to the economic benefits of farmers[2]. The grain loss mainly includes header loss, unthreshed loss, separating loss, cleaning loss and leakage loss. The header loss refers to the falling of crops caused by header vibration and collision, which varies with the natural attributes of crops (maturity, lodging, moisture content, etc.); the unthreshed loss refers to the loss caused when the seeds are discharged from the straw outlet of combine harvester without being removed during the threshing process; the separating loss refers to the loss caused when the seeds are entrained in the stalks and impurities during threshing and discharged from the straw discharge outlet without entering the cleaning device; the cleaning loss refers to the loss caused when the grains and impurities are discharged or blown out of the machine together during the screening and airflow separation process; the leakage loss refers to the loss caused by the leakage of grains due to poor mechanical sealing in the harvesting process[3-5].

The header loss and leakage loss are difficult to be monitored accurately in real time; the unthreshed loss is closely related to the structure of threshing device and accounts for a small proportion of the total loss; the separating loss and cleaning loss are mainly affected by controllable factors such as feeding amount and cleaning device working parameters, and they account for a large proportion of the total loss, directly affecting the operation quality of grain combine harvester[6-10]. Therefore, the researches on grain loss monitoring methods and devices at home and abroad focus on the separating loss and cleaning loss of grain combine harvester. What’s more, the monitoring methods and devices for separating and cleaning loss sensors are more developed in developed countries in Europe and America, and they have been sold as accessories on large grain combine harvesters. In China, cleaning loss sensors developed by some scientific research institutions have reliable performance and are close to real-life application. However, the development of the separating loss sensors is still in the theoretical and experimental stage, which is far behind the foreign technical level.

Accurate real-time grain loss information can not only provide intuitive basis for the machine operator to adjust the working parameters of combine harvester but also ensure the operation quality and efficiency. It is convenient for farmers and relevant government departments to track and monitor harvest losses in various regions. In addition, it can provide feedback of the operation quality for adaptive parameter adjustment system and automatic driving system of combine harvester. In order to realize these, it is necessary to have high precision and reliable separating and cleaning loss sensors. However, the monitoring methods and devices of separating loss are relatively underdeveloped, making it an urgent problem in the development of grain combine information intelligence in China. Thus, in this paper, the monitoring principles of the separating loss, the sensor signal processing system, the sensor damping method, the probability distribution model of separating loss and other aspects are analyzed. The research status of the monitoring methods and devices of the separating loss of grain combine harvester at home and abroad are summarized, and the future development direction is prospected. This could promote continuous improving and upgrading of the loss monitoring sensors in China and meet the needs of modernization requirements for harvesting operations and precision agriculture.

1 Monitoring principles of separating loss

1.1 Typical monitoring principles

As early as the 1960s, scholars from developed countries in Europe and America began to explore the principles of separating loss monitoring of combine harvesters, and large foreign agricultural machinery enterprises deepened their research and obtained many research results and test data. At present, the separating loss monitoring sensor has been widely used in foreign advanced combine harvesters The JD9660STS combine harvester produced by JOHN DEERE company from the United States, LEXION760 combine harvester produced by CLAAS company from Germany, etc. have all been equipped with information monitoring system including separating loss monitoring sensor[11-12]. Some foreign technology companies have also developed some separating loss monitoring devices according to the needs of users and sold in the market. In the 1990s, scholars in China began studying the monitoring principles of separating loss, and Guangyun Qi, Xinzhong Wang, etc. successively put forward the monitoring scheme[13-14]. In the past 10 years, with further research on loss monitoring principles by Chinese scholars, the monitoring accuracy and stability have been significantly improved. After sorting out and classifying the relevant literature at home and abroad, the typical monitoring principles of separating loss are shown in Table 1.

Tab. 1 Typical monitoring principles of separating loss

It can be seen from Table 1 that the separating loss monitoring method is mainly based on the principles of vision, piezoresistance and piezoelectricity. The application of the vision sensor is limited by the working environment of the combine harvester, and its effect is not ideal; the piezoresistive sensor needs additional power supply compared with the piezoelectric sensor, and its response sensitivity to instantaneous impact force is low; based on the piezoelectricity principle, materials such as grains, stalks and impurities impact the separating loss sensor, and the piezoelectricity materials produce voltage signals with different frequency and amplitude to distinguish different materials. It does not need extra power supply, and can generate electric potential difference when it is impacted or pressed weakly, and it also has large response frequency bandwidth, high sensitivity, good dynamic characteristics, simple structure and other advantages[20]. It is the dominant method in the research and practical application of separating loss monitoring sensor.

1.2 Research on the monitoring principle of separating loss based on piezoelectric crystal/piezoelectric ceramic

The piezoelectric crystal generally refers to piezoelectric single crystal, which has no symmetrical center but has piezoelectric properties, such as SiO2, LiNbO3, etc.; the piezoelectric ceramics are polycrystalline piezoelectric materials, which are synthesized by artificial preparation. But before 1942, piezoelectric materials only included piezoelectric single crystals. From 1942 to 1945, BaTiO3was discovered and was categorized as the new piezoelectric material[21]. In 1954, Jaffe[22], from the United States, found the PZT performs better than BaTiO3, which opened up new possibility of applications of piezoelectric ceramics. Compared with other piezoelectric crystals or ceramics, the PZT has advantages such as higher piezoelectric coefficient and dielectric constant, and is the first choice of piezoelectric sensors at present.

In 1985, Australian scholars Eldredge et al.[23]took the lead in using piezoelectric ceramics as the sensitive element of the grain separating loss monitoring sensor. The detection of material impact force was realized by pasting piezoelectric ceramics on the back of the sensitive plate. In 1991, American scholars Strubbe et al.[24]also proposed similar loss monitoring methods, and designed circuits that can effectively distinguish grain from other materials and reduce sensor resonance. Not long later, many multinational agricultural machinery enterprises began to study this method. Till now, most researches on separating loss monitoring methods in foreign countries have been carried out in companies, and the related products are relatively mature, which have been sold as an option of grain combine harvester.

In 2006, Junfeng Li[18]of Henan University of science and technology first proposed the design scheme of grain loss sensor based on piezoelectric ceramics in China. It used a stainless steel plate with a groove and a piezoelectric ceramic plate, which met the test requirements with a good damping and vibration isolation structure design. Later, domestic scholars began to study this monitoring method. In 2008, Hanping Mao’s team[25]of Jiangsu University used ANSYS software to analyze the stress of the sensitive plate of loss sensors, determined the number and location of piezoelectric crystal installation, and verified the accuracy of the theoretical analysis through the grain impact test; In 2015, Yaoming Li’s team[26]of Jiangsu University conducted a certain research on the monitoring sensor of separating loss based on piezoelectric ceramics, determined the installation position of the sensor through the distribution test of threshed mixture, installed the sensor under the threshing roller (Figure 1) to indirectly measure the separating loss, and conducted preliminary test verification.

(a) Left view

(b) Front view

In the study of the monitoring principles of the separating loss based on piezoelectric crystal/piezoelectric ceramic, the detection of the material impact signal is realized by pasting piezoelectric crystal/piezoelectric ceramic on the sensitive plate. The method is simple and reliable with low cost and can meet the needs of real-time monitoring of different weak impact signals. However, this method is limited by the attenuation factor of the damping structure of the sensitive plate, and the harmonic vibration of the sensitive plate will impact the detection accuracy and resolution of the loss detection system. More in-depth research needs to be done in the aspects of sensor damping, sensitive plate material, damping structure, etc.

1.3 Research on the monitoring principle of separating loss based on piezoelectric polymer

Piezoelectric polymer refers to PVDF (polyvinylidene fluoride) film and other piezoelectric film materials. In the middle of the 20th century, some scholars began to study the piezoelectric polymer, and began to synthesize piezoelectric polymers artificially. However, their piezoelectricity is very low, thus they have no practical value[27]. Until 1969, Kawai[28]from Japan reported that PVDF materials can produce relatively high piezoelectricity after being polarized under high temperature and high voltage, which has industrial application value, thus marking the historic turning point of the researches of piezoelectric polymers. With the development of science and technology and the emergence of various new manufacturing processes, the piezoelectric properties of PVDF are greatly improved, and the application scope of PVDF is expanded. This kind of material has the advantages of good flexibility, low density, low impedance and high piezoelectric stress constant. It has been developed rapidly and applied widely.

In 2008—2009, Junfeng Li et al.[19-29]of Handan University discussed the feasibility and advantages of PVDF piezoelectric film in the application of grain loss sensor, and determined the main process of making grain loss sensor by PVDF piezoelectric film based on existing research results. In 2010, Liming Zhou et al.[30]of Chinese Academy of Agricultural Mechanization Sciences designed an array grain loss monitoring system with PVDF piezoelectric film as the sensing element, and calibrated the developed grain loss monitoring sensor by selecting rice grains with different moisture content. From 2011 to 2012, Zhan Zhao et al.[31-32]of Jiangsu University studied the separating loss sensor based on PVDF piezoelectric film, designed a new PVDF piezoelectric film installation structure, as shown in Figure 2, and installed it under the threshing roller to monitor the separating loss. In 2018, German scholars Bischoff et al.[33]invented a sensor that can detect the position of grain separating collision. The sensor’s sensitive elements are divided into two layers, the upper layer is PVDF piezoelectric film with multiple isolated X-direction area arrays and the lower layer is Y-direction array. This method can distinguish the simultaneous impact signal of grain and reduce the loss monitoring error. Similar to this method, Ying Sun[10]and Rui Cao[34]from Zhejiang University carried out researches separating loss sensor based on double-layer cross PVDF piezoelectric film in 2018 and 2019, respectively. Furthermore, based on the multi-sensor fusion theory, the relationship between the response characteristics of the sensor to the material collision signal and the response of the upper and lower sensing units is analyzed, and the experiments showed that this type of sensor can identify the continuous and simultaneous collision of multiple grains.

Fig. 2 Installation structure of PVDF piezoelectric film1. bases 2. lower isolators 3. figid floating raft4. upper isolatirs 5. upper rigid panels 6. PVDF films

In current researches, PVDF piezoelectric film or the piezoelectric film with PET plastic film attached on its surface is directly used as the sensitive plate, which can not only give full play to the advantages of large sensing area but can also effectively suppress the vibration and noise of traditional sensitive plate metal and improve the monitoring accuracy. However, in field operation, the moisture or dust will be adsorbed on the film, which will affect the reliability of the loss monitoring sensor. In addition, the cost per unit area of PVDF piezoelectric film is higher than that of piezoelectric ceramic, thus it is necessary to use the PVDF piezoelectric film depending on actual situations.

2 Research progress of sensor signal processing system

The separating loss monitoring sensors based on vision and piezoresistive principle are limited and have little research or practical value. Therefore, the review of sensor signal processing system in this paper is mainly focused on piezoelectric grain loss sensor. The essence of piezoelectric sensor is to realize the conversion interface from pressure excitation signal to electric signal. More specifically, it needs to cooperate with the signal processing system to transform and process its output signal, and extract the effective information carried in the signal, so as to provide accurate real-time data for the realization of human-computer interaction function.

In 2006—2007, Korean scholars Choi et al.[35-36]designed the charge amplifier, reverse amplifier, 60 Hz notch filter and reverse amplifier for adjusting the output voltage signal of PVDF piezoelectric film sensor. The signal is collected by C8051F311 single chip on-chip analog-to-4digital converter, and the collected data is transmitted to the upper computer through RS232 serial communication. In order to solve the problems of strong vibration of combine harvester and the weak signal of cleaning loss grain, Hanping Mao’s team[37]of Jiangsu University proposed to apply Butterworth filter to separate vibration noise and grain signal characteristics of unit in 2010. The order of filter was calculated according to the width of signal transition band and the attenuation of stopband, and the design of the filter was completed through simulation and test. The results show that the filter can extract the signal of grain from the strong background machine noise. In 2014, French scholars Boukabache et al.[38]designed the signal processing system of piezoelectric ceramic sensor. The different input mode charge amplifier and dual T 50 Hz notch filter (Figure 3) were used to adjust the output signal of the sensor. The collected data is transmitted to the upper computer through the universal serial bus interface. The wavelet decomposition algorithm was used in the upper computer to complete the feature extraction of the collected signal. In 2016, Indian scholars Mothe et al.[39]designed the signal processing system of piezoelectric sensor with DSP as the core controller. The signal conditioning part is mainly composed of a 10-time gain preamplifier circuit and a third-order Butterworth band-pass filter with a passband frequency of 20 kHz～80 kHz. As a result of the combination of DSP and wavelet transform algorithm, the background noise in the effective signal can be further filtered out.

Fig. 3 Schematics of double T 50 Hz notch filter topology

Based on the review of the research progress of the signal processing system of piezoelectric sensors at home and abroad, it can be seen that the output charge signal of piezoelectric materials under the external excitation is weak and lossy, so the charge amplifier is generally used as the front-end interface conditioning circuit of the sensor; in harvesting environment, the output voltage signal of the charge amplifier also includes external interference such as mechanical vibration. Therefore, band-pass filtering, high pass filtering and other methods are needed to filter external interference. The selection of filtering frequencies and methods is also the key to distinguish the seeds from other materials. The waveform must then be shaped and compared to form a rectangular square wave which is easy to count. Most of the existing filtering methods of separating loss sensor are traditional. In future researches, we can explore new hardware filtering or combine new software filtering algorithm to further improve the adaptability and versatility of the separating loss sensor in complex environment.

3 Research progress of sensor damping method

In order to reduce the difficulty of collision signal filtering and improve the monitoring accuracy of the sensor, relevant scholars have made a series of researches and tests on the damping method of the separating loss sensor, as shown in Table 2.

Tab. 2 Vibration reduction method of separating loss sensor

The vibration and noise caused by the movement of different parts of the machine (rotation of threshing roller, vibration of cleaning screen, etc.) and the impact of the ground turbulence during the working process of the combine harvester are large The signal generated by the combine harvester will greatly interfere with the output signal of piezoelectric materials, resulting in the increase of measurement error. Therefore, it is necessary to reduce the vibration of the mounting structure of the separating loss sensor. In addition, for the separating loss sensors whose sensitive element are piezoelectric ceramic, the grain is indirectly colliding with the sensitive element and the sensitive plate will produce obvious periodic attenuation oscillation signal. If the response time is long, the front and rear impact signals will overlap when multiple grains collide with the sensitive plate at the same time. In order to reduce the impact of the collision signal superposition, it is necessary to delay counting pulse, which can be distinguished in unit time. Therefore, it is necessary to increase the damping ratio of the vibration system to quickly attenuate the harmonic vibration of the sensitive plate, which can increase the detection frequency and accuracy of the sensor.

The research on the vibration reduction method of the separating loss sensor is mainly realized by optimizing the structure of the sensitive element or the sensitive plate and adding the vibration reduction/isolation/damping materials, etc., which has achieved good vibration reduction effect. However, a series of materials currently used are traditional stainless steel, rubber, etc. In future researches, we can continue to conduct in-depth research on new sensitive plate materials, vibration reduction/isolation/damping or sensor installation methods to achieve better damping effect.

4 Research progress of probability distribution model of separating loss

The real-time monitoring of the separating loss rate needs to detect the number of particles in the mixture through the separation loss sensor, then calculate the total real-time grain loss according to the probability distribution model of separating loss, and finally calculate the grain loss rate combined with the real-time yield of the combine harvester. The accuracy of probability distribution model of separating loss directly affects the accuracy of total harvest loss and real-time grain loss rate. In addition, it has a guiding role in the size and installation position of the separating loss monitoring sensor. Therefore, scholars around the world have conducted a series of probability distribution models of separating loss for different types of threshing devices.

In 2003 and 2008, German scholar Kutzbach[48]and US scholar Miu[49-50]proposed the calculation method of threshing separation model and the threshing probability model of axial and tangential threshing devices, including the calculation formula of the distribution of unthreshed, entrained and separated grains in the axial direction of threshing space. In 2008, Shujuan Yi et al.[51]of Heilongjiang Bayi Agricultural University conducted a trial study on the distribution of grains, broken grains, light impurities and other materials in the products of the screw blade plate tooth combined type and the nail tooth type axial flow threshing and separating device developed by themselves, and obtained the axial and circumferential distribution curves and equations of different products of the two devices. In 2010, Yaoming Li’s team[52]of Jiangsu University studied the distribution law of the rice, determined the most suitable location for installing the monitoring sensor of the separating loss according to the distribution of the grain, and established the mathematical model for the monitoring of the separating loss. In 2012—2013, the team established the probability model of threshing and separation for rice, wheat tangential and longitudinal axial flow cylinder respectively, and selected the position with the most stable and the smallest change in proportion of the longitudinal and transverse distribution of grain and impurity mass (Figure 8), and then compared the falling speed of grains and impurity in different positions during the falling process, the points with large difference in voltage signal and small resonance effect caused by the impact of the array PVDF separating loss sensor were used as the installation positions of the sensor[53-56].

At present, the research on the progress of probability distribution model of separating loss is not comprehensive enough most of them are tests under specific working conditions or a small number of changing working conditions. However, in actual harvest environment, there are many factors that affect the ratio of sensor monitoring loss to total loss, and a large number of tests or algorithms are needed to modify the formula. In addition, in the existing research, the grain yield in the formula of real-time grain loss rate is generally estimated using parameters such as the forward speed of the machine and the per mu yield of the crop. In future researches, the monitoring sensor of separating loss and cleaning loss can be combined with the real-time yield sensor to get more accurate real-time grain loss rate.

(a) Left view

(b) Front view

5 Development trend analysis

1) The development moves towards universality. On the basis of purchasing a general-purpose combine harvester, users can choose a few components or assemblies which can be changed and disassembled to adapt to the harvest operation of different crops, and the matching separating loss monitoring sensor also needs to be general. When the existing separating loss monitoring device is used for different crops, it is necessary to redesign the signal processing circuit, test calibration, etc. By studying the general-purpose sensor of separating loss monitoring and establishing databases of typical crop variety attributes and databases of signal processing, the same sensor can meet the monitoring needs of rice, wheat, corn, rape and other types of combine harvesters to reduce the production, sales and use costs.

2) The development moves towards information fusion. The separating loss information can be integrated with cleaning loss, real-time yield, operation route and other information to generate real-time loss distribution and yield distribution map, to further realize remote monitoring, to provide accurate information of grain loss and yield in each region for relevant departments, and to provide real-time working status and maintenance tips of cleaning device for enterprises, drivers and farmers. The separating loss information can also be linked with the adaptive control system of the working parameters of the harvester (such as the speed of threshing cylinder, the length of threshing nail teeth, etc.). The automatic adjustment of the working parameters is realized to keep the working quality in good range at all times, which can improve the working efficiency and reduce the labor intensity.