JP3144818B2 - Grain quantity detector in threshing equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a threshing apparatus in a general-purpose combine harvester for harvesting rice, wheat, soybeans, etc., in which threshed grain culms are threshed, and then the treated straw and the like are discharged. The present invention relates to a device for detecting the amount of grains mixed in a culm when the culm is released outside the machine.

[0002]

2. Description of the Related Art As disclosed in Japanese Patent Application Laid-Open No. Sho 62-169940, a general-purpose combine for harvesting rice, wheat, soybeans, etc., has a screw-type handling cylinder. In addition, there is provided a pre-cutting treatment device for cutting and introducing the planted grain stem from the root side. A receiving net (cone cave) is stretched below the handling room, and a sorting mechanism for performing swing sorting and wind sorting is provided below the receiving net. Further, at the rear of the sorting mechanism, a culm processing chamber having a built-in culm processing device is provided in communication with the culm processing apparatus, and the entire amount of the cut culm is introduced into the chamber and threshed by the screw-type cylinder. Thing that has been threshed and the grains and straw debris have fallen through the receiving net is subjected to the rocking sorting action and the wind sorting action by the sorting mechanism. While the waste such as straw waste is sent to the culm processing chamber by being put on the wind of the sorting fan in the sorting mechanism, the straw waste and the culm falling from the end of the receiving net at the rear of the handling cylinder, The culm processing device in the culm processing chamber is configured to discharge the culm and the like outside the machine. In addition, as shown in Japanese Patent Application Laid-Open No. 62-263418, in a normal threshing apparatus such as a combine, although the shape of the teeth protruding on the outer periphery of the handling cylinder is different, other basic configurations are substantially the same. .

[0003] In this case, the culm falling from the end of the receiving net is often in the form of a lump, into which grains called "sari" are mixed. If such culms are discharged outside the machine as it is, useful grains will be lost. This loss is usually called No. 4 loss. In order to reduce the fourth loss, the culm falling from the end of the receiving net is shaken while being received at the sieving line portion in the swing sorting mechanism, and is subjected to the wind sorting action by the sorting wind to reduce the grains. I try to lead it to the number reduction gutter. Accordingly, a piezoelectric sensor is disposed at a position behind the end of the receiving net, and the piezoelectric sensor detects both the falling kernels and the culms. However, by simply integrating the detection values of the piezoelectric sensor,
There was a problem that the amount of grain as the fourth loss could not be accurately detected.

In Japanese Patent Application Laid-Open No. Sho 62-263418, the size of the vortex flow rate of air generated when the grains falling at the sieve line are subjected to the sorting wind is determined by the superposition of the culm processing chamber. It has been proposed to detect the second reduction target amount indirectly by detecting with a sound wave sensor.

[0005]

However, even in the case of this indirect detection, since the culm on the culm coming into the culm processing chamber is on the culm, the culm flowing with the culm is also superfluous. It was impossible to solve the problem that it was detected by the sound wave sensor and that the grain amount could not be accurately detected. The present invention focuses on the fact that the detection waveform pattern of the kernel by the piezoelectric sensor and the detection waveform pattern by the culm are dissimilar, and processes the detection result of the piezoelectric sensor using a new data processing method called a neural network. It is an object of the present invention to solve the above-mentioned problem with a control means that can be used.

[0006]

In order to achieve this object, the present invention provides a receiving net below a handling cylinder for threshing a harvested grain culm, and a sorting mechanism below the receiving net. In a threshing device configured to discharge the culm from the rear culm outlet to the outside of the machine, in the threshing device, between the receiving net below the rear-side portion of the handling cylinder and the sorting mechanism and the sorting mechanism, the threshing leaks. Provide a piezoelectric sensor to detect the object,
A neural network is constructed by using the detection waveform of only the kernel by the piezoelectric sensor and the detection waveform of only the culm as learning data, and the amount of kernel leakage is detected from the detection waveform of the object leaking from the receiving net. Control means for performing the operation.

[0007]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, an embodiment applied to a general-purpose combiner will be described.
A machine mounted on the truck frame 3 with a reference numeral, 4 is a threshing unit provided on one side of the machine 1, 5 is a grain tank for taking out grains through a fryer, 8 is an engine provided at the rear of the grain tank 5, Reference numeral 9 denotes a cutting pre-processing unit which is attached to the front of the machine base 3 via a hydraulic cylinder 10 so as to be able to move up and down, 11 denotes an operating unit cabin located in front of the grain tank 5, and 12 denotes grains in the grain tongue 5. This is an unloading auger that takes out of the machine.

The cutting pre-processing section 9 is opened at the front of the threshing section 4 and has a rectangular tubular feeder house 13 which can be moved up and down by a hydraulic cylinder 10, and a horizontally long bucket-shaped platform which is attached to the front end of the feeder house 13. 14 and a grain augmenting auger 15 which is disposed in the platform 14 and rotates in the direction of the arrow in FIG. 1, and is capable of moving up and down at the front upper portion thereof and rotating and driving in the direction of the arrow to scrape the grain. Reel 16
And a clipper-like cutting blade 17 laterally provided at the front end of the lower surface of the platform 14 for cutting the root portion of the planted grain culm. It is collected in the central part and sent to the threshing unit 4 by the supply chain conveyor 18 in the feeder house 13.

Reference numeral 19 denotes a rotating shaft 21 extending longitudinally in the handling room 20 of the threshing unit 4 along the traveling direction of the combine.
A screw blade 22 serving as a screw-type toothing is wound around the outer periphery of the body of the handling cylinder 19, and a stem-cum-scattering plate 23 is provided at the rear end of the outer circumference of the handling cylinder 19. Is established.

Reference numeral 32 denotes a handling chamber 2 below the handling cylinder 19.
The reference numeral 25 designates a sorting mechanism provided below the handling room 20, with the receiving net being stretched while leaving the exhaust port 24 at the rear end of 0. The sorting mechanism 25 includes a swinging plate 28 that swings in the front-rear direction by swinging links 26 and 27, and a feed pan 29 located at a lower front portion of the handling cylinder 19, and a feed pan 29 provided behind the feed pan 29. One chaff sheave 30, a sorting net 31 stretched underneath, and a first chaff sheave 30
A sieve (screen) line 33, a second chaff sheave 34 provided below and below the first sieve line 33, a second sieve line 35,
The rocking board 28 is provided with the first and second flowing grain plates 36 and 37.

Reference numeral 38 denotes a dust removal fan for supplying a sorting wind to the upper surface side of the feed pan 29, and reference numeral 39 denotes the first tray 3.
6 is a first sorting fan which supplies the sorting wind toward the culm processing chamber 42, which will be described later, and which is a first sorting fan. The first conveyor with a gutter 41, reference numeral 43 is the second chaff sheave 3
And supplying the sorting wind toward the culm processing chamber 4 described later.
A second sorting fan that sends a sorting wind toward
The sorting fan 43 is provided at a position closer to the culm processing chamber 42 than the first sorting fan 39 is.

Reference numeral 44 denotes a second conveyor with a second gutter 45 for sending a second processed material, which is a reductant such as a grain from the second flow grain plate 37, to a second reduction cylinder 46. Cylinder 46
The upper end (discharge port) is opened at a position near the front end of the handling cylinder 19 in the handling chamber 20 or opened on the feed pan 29 to reduce the second processed material. Dust port 2
The inclined culm plate 47 at the lower end of the culm processing chamber 42 that communicates with the rear end of the sorting mechanism 25 faces the rear end side of the rocking plate 28, and the culm processing device 42 is provided in the culm processing chamber 42. Rotating diffusion tooth 4
8 is attached to the horizontal rotating shaft 49, and the fixed teeth 50 facing the side of the rotating diffusion teeth 48 so as to overlap with the rotation trajectories of the plurality of rotating diffusion teeth 48 driven to rotate in the direction of the arrow are fixed to the culm processing chamber. It is mounted so as to be able to protrude upward and downward from the bottom plate 54 of 42.

A plurality of culm guide rods 51 each having a base end attached to the dust outlet 24 are suspended so as to overlap with the rotation locus on the side of the rotating diffusion teeth 48. The culm processing chamber 24 is disposed so as to extend downward from above at a front position. From the culm outlet 52 provided rearward below the culm processing chamber 42, the culm scraped out by the rotating diffusion teeth 48 is formed in a horizontally long shape so as to be discharged outside the machine. The first sorting fan 39 and the second sorting fan 43 are provided on the upper rear end of the culm processing chamber 42.
This is provided so as to open a discharge air passage 53 for discharging the wind from the rear of the machine base 1.

Reference numeral 55 denotes a cutting blade provided at a position from the rear portion of the outer periphery of the trunk of the handling cylinder 19, and includes a fixed blade 56 protruding into the handling chamber 20 so as to be adjacent to the rotation locus of the cutting blade 55. At the position near the rear end of the outer periphery of the handling cylinder 19, the culm sent in a lump is divided into short pieces, or the culm is loosened, and the grains (sarari grains) mixed in the culm are received from the net 32 below. Leak.

With the above-described structure, the grain stalks which have been scraped by the reel 16 in the pre-cutting unit 9 and whose root has been cut off by the cutting blade 17 can be scraped by the scraping auger 15 in the platform 14.
, And is supplied to the front end of the handling drum 19 via the supply chain conveyor 18 in the feeder house 13. Handling room 20
Grains are sequentially shed from the cereal stem by the rotation of the handling cylinder 19 inside,
Grains and straw debris that have fallen through the receiving net 32 are sorted by the sorting mechanism 25.
At the same time, the grains are collected on the first gutter 41 and the first conveyor 4 is subjected to the swing sorting action and the wind sorting action of the first sorting fan 39.
From 0, it is collected in a grain tank 5 through a frying cylinder.

The processed material having a large amount of straw waste is collected at 34 places of the second chaff sheave, which is the rear part of the sorting mechanism, and is moved backward by the swing sorting action, while being subjected to the wind sorting action by the second sorting fan 43. The straw chips and dust are blown off toward the culm processing chamber 42, and together with the exhaust air, the straw chips and dust can be smoothly discharged to the outside of the machine through the exhaust path 53 on the upper rear end of the culm processing chamber 42. .

On the other hand, the grain culm discharged from the dust discharge port 24 at the rear end of the handling cylinder 19 is received on the front surface of the culm guide rod 51, falls in the direction of the rotating diffusion teeth 48, and is scraped out by the culm disposal device 48. Since the air is discharged from the culm exit 52 to the outside of the machine, the air
The culm does not clog. In other words, light materials such as straw waste and dust are discharged out of the machine almost linearly through the air discharge passage 53 on the upper side of the culm processing chamber 42 together with the blast, and the culm which is a large waste is discharged. Since the discharge part is separated into two parts so as to correspond to the nature of the material to be discharged, such as from the top to the bottom of the culm processing chamber, the sorting wind does not flow backward in the direction of the sorting mechanism in the culm processing chamber . Therefore, the second grain board 3
7 is reduced, and as a result, there is less straw waste and dust mixed with the second processed material gathered on the second conveyor 44 from the second gutter 45, so that the second processed material is transferred to the second reduction cylinder 46. When returning to the front part of the handling cylinder 19 via the, the amount of waste discharged from the dust discharge port 24 at the rear end of the handling cylinder 19 is reduced, and the loss of the fourth cutting grain can be reduced. When returning the second processed material having a small amount of straw waste and dust to the front feed pan 29 and the first chaff sheave 30 of the rocking plate 28 via the second return cylinder 46, the first gutter 4 is used.
The mixing of small debris into the inside 1 is drastically reduced, and the third loss (loss due to grains discharged outside the machine from the culm outlet 52 at the lower part of the culm processing chamber 42) can also be reduced.

Reference numeral 57 denotes a position on the rear side of the handling cylinder 19 beyond the position where the cutting blade 55 and the fixed blade 56 are installed, and at a position below the receiving net 32 (above the chaff sheave 34). There are a plurality of piezoelectric sensors. In the embodiment, the detection surfaces of the three piezoelectric sensors 57 are arranged such that the grains leaking from the receiving net 32 easily collide with the rotation of the handling cylinder 19 (see FIG. 4).

FIG. 4 is a block diagram of the control means of the present invention. Reference numeral 58 denotes an amplifier (amplifier) for amplifying an analog input signal of each of the piezoelectric sensors, and reference numeral 59 denotes a neural network process which will be described later. Reference numeral 60 denotes a read-only memory (ROM) for preliminarily storing a control program and an initial value required for arithmetic processing by the central control device 59;
Reference numeral 61 denotes a readable / writable memory (RAM) for temporarily storing various data used for arithmetic processing,
Reference numeral 62 indicates an input / output (I / O) interface unit. Note that the actuator 63 represented by an electromagnetic solenoid driven in accordance with a command signal from the central control device 59 executes a speed change of a threshing unit such as the rotation speed of the handling cylinder 19 or changes a traveling speed of the traveling machine 1. Change the threshing speed.

Next, a method for processing the signal data detected by the piezoelectric sensor 57 will be described. According to an experiment, only an appropriate amount of grain is leaked from the receiving net 32 at a position near the rear of the rotating handling drum 19. FIG. 5 shows an example of a detection waveform pattern in which the leaking kernel collides with the inspection surface of the piezoelectric sensor 57, in which the horizontal axis represents time t and the vertical axis represents the detection voltage. Similarly, FIG. 6 is an example of a detected waveform pattern when only the culm collides with the piezoelectric sensor 57.

The m experimental pattern data of only the kernel, K1, K2, K3, ‥‥, Ki, の み, Km, and the m experimental pattern data of only the culm, H1, H2, H
3, ‥‥, Hj, ‥‥ Hm are prepared. In each of the experimental pattern data, for example, in the case of only the grain in FIG.
Sampling for a time (for example, 150 milliseconds) is performed, and the sampling time is divided by n (for example, 100), and Vki1, Vki2, Vki3, ‥‥, Vkii, ‥‥, Vkin
The voltage signals are sampled. For culms, Vh
j1, Vhj2, Vhj3, ‥‥, Vhji, ‥‥, and Vhjn voltage signals.

FIG. 7 is a main flowchart for detecting the amount of grain per unit time as the fourth loss. Following start and initialization, in step 701, m pieces of experimental data of only the grain are read. , Its experimental data K
Sampling value Vki for each T / n time interval for i
Is stored (step 702). Similarly, step 703
In step 704, m pieces of experimental data on only the culm are read, and the sampling value Vhj for each T / n time interval is stored for each of the experimental data Hj. Next, in step 705, a sampling value of the experimental data of the kernel alone is input and learned by a neural network. Similarly, in step 706, a sampling value of the experimental data of only the culm is input and learned by the neural network. Next, at step 707, the learned offset value and coupling coefficient of the neural network are stored in the RAM. The detection signal of the piezoelectric sensor 57 during the actual threshing process is input to these learned neural networks, and the appearance frequency of the grain pattern is measured (step 708). Step 7
In step 09, the amount of grain per unit time as the fourth loss is calculated from the ratio between the area of the detection surface of the piezoelectric sensor 57 and the actual leakage area.

Next, the learning by the back propagation neural network will be described. FIG. 8 shows a subroutine flowchart of the neural network. The back propagation learning method in the present embodiment is a supervised learning method, and is performed on a neural network having a three-layered hierarchical structure consisting of an input layer I → intermediate layer → output layer.

In the present embodiment, the number of input layer units P in the neural network structure is such that n = 1 for each of the experimental data for the kernel only and the experimental data for the culm only.
00, the number of intermediate layer units Q is 25, and the number of output layer units S is 2. FIG. 9 illustrates the structure of the neural network. The learning by the neural network is performed only for the kernel and only for the culm, and is performed for each of the m pieces of experimental data, and the modified example of the pattern for the kernel and the modified example of the pattern for the culm are sufficiently learned. Shall be.

A specific learning method is executed as follows. First, n sample values Vki1, Vki2, Vki3, ‥‥, Vkii, ‥‥, and Vkin in the i-th experimental data of the kernel only
(Please note that among the symbols in the following text, ｛
Those surrounded by｝ indicate a plurality of units). [1] Coupling coefficients {WPQ}, {WQS} and offset values {θP}, {θQ} that determine the state of the neural network
Are initialized with small random values. [2] Sample values Vki1, Vki2, Vki3, ‥‥, Vkii, ‥
‥, Vkin is the first learning data. [3] Set the value of the learning data to the output ｛Pou of the input layer unit P
t}, the input value Qin to the intermediate layer unit is obtained using the coupling coefficient {WPQ} from the input layer to the intermediate layer and the offset value θP of the intermediate layer unit Q. This input value Qin
Then, the output value Qout of the intermediate layer unit Q is obtained by the sigmoid function f (X) such as f (X) = 1 / (1 + exp (−X)). The output Pout of the input layer unit P
Is: Pout = 1 / (1 + exp (− (X−θ) / T)} (5) where X = W1 * Vki1 + W2 * Vki2 + 2 + Wi * Vkii + ‥‥ + Wn * Vkin W1, W2, ‥‥ Wi, ‥‥, and Wn are weighted by coupling coefficients.
θ is an offset value changed by learning, and T is
A temperature parameter of an energy function by a Boltzmann machine. In the embodiment, T = 1. [4] Using the output value {Qout} of the intermediate layer unit, the coupling coefficient {WQS} from the intermediate layer to the output layer S, and the offset value θQ of the output layer Q, each output of the output layer units S1 and S2 Find the value Sout. Output Qout of the middle layer unit Q
Is derived from the same calculation formula, where Pout is X in the above formula (5). [5] The teacher signal in learning only the kernel is set to S1out = 1 for the output layer unit S1, and S2out = 0 for the output layer unit S2. For each output layer unit, from the difference between the teacher signal and the output value of the output layer,
An error δs1 with respect to the coupling coefficient connected to the output layer unit S1 and the offset of the intermediate layer unit Q is obtained. The error δs2 is similarly obtained for the output layer unit S2. [6] Error δs1 and coupling coefficient from intermediate layer to output layer ｛WQS
From｝ and the output value Qout of the intermediate layer, the intermediate layer unit Q
Coefficient and intermediate layer unit offset value θ
Find the error σq for P. [7] Error s1 in output layer unit S1 obtained in [5]
And the product of the output Qout of the intermediate layer unit Q and the constant α, the intermediate layer unit Q outputs the output layer unit S1.
Correct the coupling coefficient WQS that leads to. Further, the offset value θQ of the output layer unit S1 is corrected by adding the product of the error σs1 and the constant β. Similarly, the output layer unit S
Also ask for 2. [8] By adding the product of the error σq in the intermediate layer unit Q, the output value Pout of the input layer unit P, and the constant α,
Correct the coupling coefficient WPQ from the input layer unit P to the intermediate layer unit Q. Also, by adding the product of the error σq and the constant β, the offset value θP
To correct.

[9] Input the next learning data. [10] Return to [3] until the learning data ends. [11] The number of learning repetitions is updated. [12] If the number of learning repetitions is equal to or less than the limit number,
Return to [2].

In the final input layer unit P, the output value Pout is compared with the teacher signal, the error σs1 and the error σq are back-propagated, the weight is corrected, and learning is performed. Learning is terminated when the difference between the output value Pout of the input layer unit and the output Sout of the output layer unit becomes equal to or smaller than the allowable error. The same learning as described above is performed for the experimental data of only the culm. In this case, the teacher signal for the culm in the flow [5] inverts the teacher signal for the kernel. That is, for the output layer unit S1, S1out =
0, and S2out = 1 for the output layer unit S2. As described above, at the time of learning the experimental data for the grain and the learning of the experimental data for the culm, it is possible to reliably provide the teacher signal, so that the learning result must be converged. The offset value can be reliably obtained in advance.

[0027]

As described above, according to the present invention, the pattern when the target grain collides with the piezoelectric sensor,
The difference from the pattern when the culm collides with the piezoelectric sensor can be learned in advance by the neural network. Therefore, even if an object to be detected (grain) is mixed with a foreign substance (culm) and simultaneously detected by the piezoelectric sensor, the detection pattern is the learning pattern of the kernel for which the learning has been completed. In the case similar to the above, it can be determined that the object is a grain, so that even in the actual detection, it is possible to easily detect only the object to be detected (the grain). As a result, the amount of grain leakage as the fourth loss can be detected.

[Brief description of the drawings]

FIG. 1 is a side view of a combine.

FIG. 2 is a plan view of the combine.

FIG. 3 is a sectional view of a threshing apparatus.

4 is a block diagram of a control device and an arrangement relationship of piezoelectric sensors shown in a section taken along a line IV-IV in FIG. 3;

FIG. 5 is a diagram showing a detection pattern of a grain-only piezoelectric sensor.

FIG. 6 is a diagram showing a detection pattern of only a culm by a piezoelectric sensor.

FIG. 7 is a main flowchart of the present invention.

FIG. 8 is a flowchart of a back propagation method.

FIG. 9 is a diagram showing a structure of a neural network.

[Explanation of symbols]

 Reference Signs List 4 threshing section 9 cutting pretreatment section 13 feeder house 18 supply chain conveyor 19 handling cylinder 20 handling chamber 24 exhaust port 25 sorting mechanism 28 rocking board 30 first chaff sheave 32 receiving net 34 second chaff sheave 36 first flow grain plate 37 first 2nd grade grain plate 42 culm processing room 46 2nd reduction cylinder 53 exhaust air passage 55 cutting blade 56 fixed blade 57 piezoelectric sensor 59 central controller 60 read only memory (ROM) 61 readable / writable memory (RAM) 62 input / output interface

──────────────────────────────────────────────────続 き Continued on the front page (58) Fields surveyed (Int. Cl. ⁷ , DB name) G01N 33/10 G01N 5/00-9/36 A01F 12/18 A01F 12/54 JICST file (JOIS)

Claims (1)

(57) [Claims] 1. A receiving net is provided below a handling drum for threshing a harvested grain culm, and a sorting mechanism is provided below the handling net, and the discharged culm is discharged outside the machine from a discharging culm outlet behind the sorting mechanism. In the threshing apparatus, a piezoelectric sensor for detecting an object leaking from the receiving net is provided between the receiving net below the rear part of the handling cylinder and the sorting mechanism; A control for constructing a neural network using the detection waveform of only the kernel by the sensor and the detection waveform of only the culm as learning data, and detecting the amount of grain leakage from the detection waveform of the detected object leaking from the receiving net. A grain amount detecting device in a threshing device, characterized by comprising means.