Paddy field cultivator and paddy field tillage depth detection device and method thereof

Abstract

A paddy field cultivator and a paddy field plowing depth detection device and method thereof, the paddy field cultivator comprising the paddy field plowing depth detection device, the device comprising: a ground profiling mechanism for detecting the depth of the mud shoe sinking and the change data of the terrain ups and downs when the traction machine is working; an inclination sensor for detecting the three-point suspension inclination change of the traction machine to indirectly obtain the plowing depth data; a machine body gyroscope for detecting the roll and pitch attitude of the traction machine to correct the plowing depth detection error caused by the attitude transformation when the traction machine is working; a BDS double-antenna system for detecting the absolute value of the working track and the plowing depth change of the traction machine; and a data processor in communication connection with the profiling mechanism, the BDS double-antenna system, the machine body gyroscope and the inclination sensor respectively, receiving and saving the data collected by the profiling mechanism, the BDS double-antenna system, the machine body gyroscope and the inclination sensor. The application also provides a paddy field plowing depth detection method.

Description

Technical Field

[0001] This invention relates to agricultural intelligent detection and information fusion technology, and in particular to a paddy field tillage machine and a paddy field tillage depth detection device and method that utilizes multi-sensor information fusion to detect rotary tillage in paddy fields. Background Technology

[0002] Paddy field tillage is a crucial step in rice production. Precise tillage improves water and fertilizer utilization, irrigation uniformity, reduces weeds and pests, and mitigates environmental pollution. Tillage depth is the most important parameter for evaluating the quality of paddy field tillage and has significant implications for crop growth and yield. Insufficient tillage depth fails to meet agricultural technical requirements, hinders root growth, and negatively impacts crop development and yield; excessive tillage depth increases tractor power consumption, raises operating costs, and reduces operational efficiency. Therefore, accurately measuring the tillage depth of rotary tillage in paddy fields is of paramount importance.

[0003] Current rotary tillage depth detection technologies generally employ a single tilt sensor, mounted on the cantilever arm of the tractor and rotary tiller connected by a three-point suspension. This sensor detects changes in the cantilever arm's tilt angle and indirectly calculates the tillage depth by establishing a geometric model. While this method is recognized for its accuracy and reliability in dryland tillage, it becomes less effective in scenarios involving dry rotary tillage in paddy fields with complex terrain and high soil viscosity. The significant changes in the tractor's footing within the mud during operation lead to inaccurate calibration of the tilt sensor, resulting in substantial errors in the detected tillage depth. Furthermore, the geometric model established in current technologies is simplified to reduce computational load, assuming the agricultural machinery operates on a flat, hard surface and ignoring the impact of terrain variations on the tillage depth detection data. However, in the dry rotary tillage operation in paddy fields, the tractor undergoes roll and pitch changes, rendering the simplified model unsuitable for the specific working conditions and leading to significant errors in the detection data. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to address the above-mentioned deficiencies of the prior art by providing a paddy field tillage machine and a paddy field tillage depth detection device and method that utilizes multi-sensor information fusion to detect the tillage depth of paddy field rotary tillage.

[0005] To achieve the above objectives, the present invention provides a paddy field tillage depth detection device, comprising:

[0006] The ground contouring mechanism is used to detect changes in the depth of the tractor's foot sinking into the mud and the terrain's elevation changes during tractor operation.

[0007] An inclination sensor is used to detect changes in the inclination angle of the three-point suspension of the tractor to indirectly obtain tillage depth data;

[0008] The fuselage gyroscope is used to detect the roll and pitch attitude of the tractor, so as to correct the tillage depth detection error caused by the attitude change of the tractor during operation.

[0009] The BDS dual-antenna system is used to detect the absolute value of the tractor's operating trajectory and tillage depth changes; and

[0010] The data processor is communicatively connected to the contouring mechanism, the BDS dual-antenna system, the fuselage gyroscope, and the tilt sensor, respectively, and receives and saves the data collected by the contouring mechanism, the BDS dual-antenna system, the fuselage gyroscope, and the tilt sensor.

[0011] In the aforementioned paddy field tillage depth detection device, the data processor uses a microcontroller development board and is connected to the contouring mechanism, BDS dual-antenna system, fuselage gyroscope, and tilt sensor via a CAN bus. The collected data is saved to an SD card using an SDIO controller.

[0012] In the aforementioned paddy field tillage depth detection device, the data processor drives the LCD screen to display the working status in real time via FSMC.

[0013] The aforementioned paddy field tillage depth detection device, wherein the ground contouring mechanism has a symmetrical structure, comprising:

[0014] Horizontal support;

[0015] The trays are symmetrically arranged on both sides of the center line of the horizontal support and are connected to the horizontal support through the connecting frame. The trays are installed at the bottom of the connecting frame and the bottom of the trays are in contact with the ground.

[0016] A limiting plate is installed on the horizontal support and is connected and cooperated with the connecting frame to realize the degree of freedom of the pallet in the horizontal rolling direction;

[0017] A contour-following gyroscope is mounted on the connecting frame; and

[0018] A return spring is mounted on the connecting frame.

[0019] In the aforementioned paddy field tillage depth detection device, the connecting frame is a rectangular frame, the front vertical rod of the connecting frame is connected to the horizontal support, and the bottom end of the rear vertical rod of the connecting frame is hinged to the support plate to ensure that the support plate is in contact with the ground during the contouring process.

[0020] In the aforementioned paddy field tillage depth detection device, the contour gyroscope is installed on the upper crossbar of the connecting frame and connected to the data processor. It is used to detect the angle change of the upper crossbar and obtain quantitative data on the sinking of the machinery and the undulation of the terrain during the contouring process.

[0021] In the aforementioned paddy field tillage depth detection device, the limiting plate is an arc-shaped plate, and both ends of the arc-shaped plate are connected to the horizontal support; the middle of the arc-shaped plate is an arc-shaped through hole, and the front vertical rod is provided with a limiting boss that matches the middle arc-shaped through hole.

[0022] To better achieve the above objectives, the present invention also provides a method for detecting the tillage depth of paddy fields, comprising the following steps:

[0023] S100. The ground contouring mechanism is installed behind the tractor to detect the ground shape and the subsidence of the implements before tillage, which serves as the zero reference value for tillage depth.

[0024] S200. Install a fuselage gyroscope directly behind the cab of the tractor to detect attitude changes during tractor operation as attitude correction values.

[0025] S300. Install the tilt sensor on the three-point suspension arm of the traction machine to detect the tilt angle change of the three-point suspension arm and obtain the indirect tillage depth value.

[0026] S400: The first and second antennas of the BDS dual-antenna system are respectively installed on the central axis of the tractor and rotary tiller roof to detect the change in tillage depth during the rotary tillage operation of the tractor.

[0027] S500. Set up an RTK base station of the BDS dual-antenna system in a non-working area and connect it to the first antenna and the second antenna respectively, so that the data calculated by the first antenna and the second antenna are the RTK corrected tillage depth change values.

[0028] S600, using the trained tillage depth detection model to correct the real-time acquired tillage depth zero-value benchmark, attitude correction value, indirect tillage depth value, and tillage depth change value; and

[0029] S700 integrates the corrected tillage depth data to obtain accurate online detection values ​​for paddy field tillage depth.

[0030] In the above-mentioned paddy field tillage depth detection method, in step S700, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the Kalman fusion algorithm is used to fuse the filtered tillage depth data; or in step S600, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the trained tillage depth detection model is used to correct the tillage depth data.

[0031] To better achieve the above objectives, the present invention also provides a paddy field tillage machine, which includes the above-mentioned paddy field tillage depth detection device.

[0032] The technical effects of this invention are as follows:

[0033] This invention is based on multi-source sensor fusion technology for detecting tillage depth in paddy field rotary tillage. It adopts a ground contouring mechanism, sensor layout and tillage depth detection model, and combines data filtering and fusion algorithms. The detected tillage depth data is corrected by using the machine sinking depth and attitude value, and the corrected data is fused to finally obtain an accurate online detection value of paddy field tillage depth, thus realizing accurate online detection of paddy field tillage depth.

[0034] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the present invention. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of a paddy field tillage machine according to an embodiment of the present invention;

[0036] Figure 2 This is a schematic diagram of a ground contouring mechanism according to an embodiment of the present invention;

[0037] Figure 3 This is a schematic diagram of the installation of a ground contouring mechanism according to an embodiment of the present invention;

[0038] Figure 4 This is a schematic diagram of a contour-following gyroscope structure according to an embodiment of the present invention;

[0039] Figure 5 This is a schematic diagram of a data processor structure according to an embodiment of the present invention;

[0040] Figure 6 This is a schematic diagram of the pitch direction movement of a traction machine or rotary tiller according to an embodiment of the present invention;

[0041] Figure 7 This is a schematic diagram of the movement of a tractor or rotary tiller in the lateral rolling direction according to an embodiment of the present invention;

[0042] Figure 8 This is a schematic diagram illustrating the establishment of the coordinate system for a detection model according to an embodiment of the present invention;

[0043] Figure 9 This is a schematic diagram of the BP-Kalman filter algorithm according to an embodiment of the present invention.

[0044] Among them, the attached reference numerals

[0045] 1 Traction machine

[0046] 11 racks

[0047] 12 Walking wheels

[0048] 2. Rotary Tiller

[0049] 3. Paddy field tillage depth detection device

[0050] 31 Ground contouring mechanism

[0051] 311 Body Connector

[0052] 312 Horizontal support

[0053] 313 Front vertical bar

[0054] 314 Rear Vertical Bar

[0055] 315 Cross Link

[0056] 316 Limit Plate

[0057] 317 pallet

[0058] 318 Return Spring

[0059] 319 Fascinating Gyroscope

[0060] 3191 package top case

[0061] 3192 PCB circuit board

[0062] 3193 Package Back Case

[0063] 32. In-body gyroscope

[0064] 33 Tilt Sensor

[0065] 34 BDS Dual Antenna System

[0066] 341 RTK base stations

[0067] 342 First Antenna

[0068] 343 Second Antenna

[0069] 35 Data Processors

[0070] 351 upper package

[0071] 352 Development Board

[0072] 353 Lower Package Detailed Implementation

[0073] The structural and working principles of the present invention will be described in detail below with reference to the accompanying drawings:

[0074] See Figure 1 , Figure 1This is a schematic diagram of a paddy field tillage machine according to an embodiment of the present invention. The paddy field tillage machine of the present invention includes a tractor 1, a paddy field tillage depth detection device 3, and a rotary tiller 2. Any self-propelled device that can provide traction power for paddy field tillage can be used as the tractor 1. The tractor 1 is preferably a tractor. The rotary tiller 2 can be any existing machine that can work in conjunction with the tractor 1, and there are no restrictions on this. The paddy field tillage depth detection device 3 includes a ground contouring mechanism 31 for detecting changes in the mud depth and terrain elevation during the operation of the tractor 1; an inclination sensor 33 for detecting changes in the three-point suspension inclination angle of the tractor 1 to indirectly obtain tillage depth data; a fuselage gyroscope 32 for detecting the roll and pitch attitude of the tractor 1 to correct tillage depth detection errors caused by attitude changes during operation; a BDS dual-antenna system 34 for detecting the operating trajectory and absolute value of tillage depth changes of the tractor 1; and a data processor 35, which is communicatively connected to the contouring mechanism, the BDS dual-antenna system 34, the fuselage gyroscope 32, and the inclination sensor 33, respectively, to receive and store the data collected by the contouring mechanism, the BDS dual-antenna system 34, the fuselage gyroscope 32, and the inclination sensor 33.

[0075] The BDS dual-antenna system 34 includes an RTK base station 341 set up in the non-operational area outside the field and a first antenna 342 and a second antenna 343 installed on the rotary tiller 2. The first antenna 342 is installed on the central axis of the top of the tractor 1, and the second antenna 343 is installed on the central axis of the rotary tiller 2. The first antenna 342 outputs the RTK differential positioning data of the tractor 1, which ultimately forms the working trajectory of the tractor 1. The first antenna 342 and the second antenna 343 jointly output the relative vertical distance between the top of the tractor 1 and the top of the rotary tiller 2, which is converted to obtain the tillage depth value. The body gyroscope 32 is installed on the central axis of the tractor 1 body, located behind the cab, and is fixed with strong double-sided tape and a belt. The tilt sensor 33 is installed on the lower cantilever of the three-point suspension and is fixed with a strong magnetic base and strong double-sided tape.

[0076] See Figure 2 and Figure 3 , Figure 2 This is a schematic diagram of the ground contouring mechanism 31 according to an embodiment of the present invention. Figure 3This is a schematic diagram of the installation of a ground contouring mechanism 31 according to an embodiment of the present invention. The ground contouring mechanism 31 of the present invention is used to detect the depth of mud sinking of the tractor 1 during operation and the condition of uncultivated terrain. The ground contouring mechanism 31 of this embodiment has a symmetrical structure and is installed on the frame 11 of the tractor 1, located directly behind the frame 11 near the rear walking wheel 12. The ground contouring mechanism 31 includes: a horizontal support 312; a support plate 317, symmetrically arranged on both sides of the center line of the horizontal support 312 and connected to the horizontal support 312 through a connecting frame, the support plate 317 being installed at the bottom end of the connecting frame, and the bottom of the support plate 317 contacting the ground; a limiting plate 316, installed on the horizontal support 312 and connected and cooperating with the connecting frame to realize the degree of freedom of the support plate 317 in the lateral rolling direction; a contouring gyroscope 319, installed on the connecting frame; and a return spring 318, installed on the connecting frame. The connecting frame is preferably a rectangular frame, including a front vertical rod 313, a rear vertical rod 314, and upper and lower horizontal connecting rods 315. The front vertical rod 313 is connected to the horizontal support 312, and the bottom end of the rear vertical rod 314 is hinged to the support plate 317 to ensure that the support plate 317 is in contact with the ground during the contouring process. The limiting plate 316 is preferably an arc-shaped plate, with both ends of the arc-shaped plate connected to the horizontal support 312. The arc-shaped plate has an arc-shaped through hole in the middle, and the front vertical rod 313 is provided with a limiting boss that matches the arc-shaped through hole in the middle.

[0077] In this embodiment, the ground contouring mechanism 31 is installed below the three-point suspension of the traction machine 1, located between the traction machine 1 and the rotary tiller 2. Two vertical fixed connecting plates with through holes are correspondingly provided at the rear end of the frame 11 of the traction machine 1. The ground contouring mechanism 31 is connected to the corresponding through holes of the fixed connecting plates via the body connector 311 using bolts and nuts. The body connector 311 and two horizontal supports 312 are fixedly connected by bolts, and two pins are symmetrically welded to the upper horizontal support 312. The support plate 317 is divided into upper and lower parts. The lower part is sled-shaped, allowing for smoother contact with the ground, while the upper part is eave-shaped, with an inclined surface that allows soil to slide off during the contouring process, preventing the support plate 317 from being compacted. The upper part of the support plate 317 is connected to the connector via bolts and nuts through the corresponding through holes. The other side of the connector and the rear vertical rod 314 are connected via bolts and nuts through corresponding connecting holes. The rear vertical rod 314 is preferably a T-shaped profile with four through holes on its web, which are used to connect to the upper horizontal connecting rod 315, the return spring 318, the lower horizontal connecting rod 315, and the connecting piece from top to bottom. The upper horizontal connecting rod 315 is preferably a long strip profile with three through holes located at both ends and the middle of the connecting rod. The lower horizontal connecting rod 315 is a long strip profile of the same length and width as the upper connecting rod, with one through hole at each end. The upper and lower horizontal connecting rods 315 are fastened to the rear vertical rod 314 at both ends by locking bolts. The front vertical rod 313 is preferably a T-shaped profile with three through holes on its web, which are used to connect to the upper horizontal connecting rod 315, the return spring 318, and the lower horizontal connecting rod 315 from top to bottom. It has a through hole on its flange that penetrates the web, located below the connection hole with the upper horizontal connecting rod and above the return spring hole. The upper and lower horizontal connecting rods 315 are clamped and fixed to the front vertical rod 313 by locking bolts. The web of the front vertical rod 313 has two cut-off grooves near the lower middle position, with a protruding portion in the middle serving as a limiting boss to cooperate with the limiting plate 316. The limiting plate 316 is preferably an arc-shaped plate with two through holes at both ends and an arc-shaped through hole in the middle. The two through holes at the ends are connected to the horizontal support 312 by bolts, and the arc-shaped through hole in the middle cooperates with the protruding portion of the web of the front vertical rod 313. The wing plate of the front vertical rod 313 is fitted into the pin on the upper horizontal support 312 and fixed to the upper horizontal support 312 by a pin.

[0078] The ground contouring mechanism 31 achieves movement in the pitch direction through the connection of the connector and the rear vertical rod 314, ensuring that the support plate 317 can better conform to the ground during the contouring process. Through the cooperation of the front vertical rod 313 and the limiting plate 316, it achieves freedom in the roll direction, ensuring that the support plate 317 can better conform to the ground during the contouring process. A contouring gyroscope 319 installed in the middle of the upper horizontal connecting rod 315 detects the angle change of the upper horizontal connecting rod 315, obtaining quantitative data on the machine's sinking and the terrain's elevation changes during the contouring process.

[0079] See Figure 4, Figure 4 This is a schematic diagram of the contour-following gyroscope 319 according to an embodiment of the present invention. The contour-following gyroscope 319 is mounted on the upper crossbar of the connecting frame and connected to the data processor 35. It is used to detect the angle change of the upper crossbar and obtain quantitative data on the sinking of the machine and the terrain undulations during the contour-following process. The contour-following gyroscope 319 includes a top encapsulation shell 3191, a bottom encapsulation shell 3193, and a PCB circuit board 3192. The bottom encapsulation shell 3193 is provided with a protrusion structure that matches the upper crossbar 315, and a through hole is opened in the center of the protrusion. The protrusion and the middle through hole of the upper crossbar 315 are fixedly installed by bolts and nuts.

[0080] See Figure 5 , Figure 5 This is a schematic diagram of a data processor structure according to an embodiment of the present invention. The data processor 35 of this embodiment can be installed in the driver's cab of the tractor 1. The data processor 35 uses a microcontroller development board 352 and is connected to the ground contouring mechanism 31, the fuselage gyroscope 32, the tilt sensor 33, and the BDS dual-antenna system 34 via a CAN bus. The data processor 35 includes an upper package 351, a development board 352, and a lower package 353. The upper package 351 has an LCD screen display hole and button holes, while the lower package 353 has an aviation connector mounting hole, a power socket, and an SD card slot. The data processor 35 preferably uses an STM32F103 series microcontroller development board. It communicates with two contour gyroscopes 319, a BDS dual-antenna system 34, a body gyroscope 32, and a tilt sensor 33 via a CAN bus. It uses an SDIO (Secure Digital Input and Output) controller to save the collected data to an SD card and drives an LCD screen via an FSMC (Flexible Static Memory Controller) to display the working status in real time.

[0081] The method for detecting the tillage depth of paddy fields according to the present invention includes the following steps:

[0082] Step S100: Install the ground contouring mechanism 31 behind the tractor 1 to detect the ground shape and the sinking of the implements before tillage as the zero reference value for tillage depth; that is, to detect the depth of the mud foot sinking when the tractor 1 is working, as well as the terrain conditions before tillage.

[0083] Step S200: Install the fuselage gyroscope 32 directly behind the cab of the tractor 1 to detect the attitude changes of the tractor 1 during operation as attitude correction values; that is, detect the roll and pitch attitude of the tractor 1 and correct the tillage depth detection error caused by the attitude change of the tractor 1 during operation.

[0084] Step S300: Install the tilt sensor 33 on the three-point suspension arm of the traction machine 1 to detect the tilt angle change of the three-point suspension arm and obtain the indirect tillage depth value.

[0085] Step S400: Install the first antenna 342 and the second antenna 343 of the BDS dual antenna system 34 on the central axis of the roof of the tractor 1 and the rotary tiller 2, respectively, to detect the change in tillage depth during the rotary tillage operation of the tractor 1; that is, use the BDS dual antenna to detect the working trajectory and the absolute value of the tillage depth change of the tractor 1.

[0086] Step S500: Set up an RTK base station 341 of the BDS dual-antenna system 34 in the non-working area and connect it to the first antenna 342 and the second antenna 343 respectively, so that the data calculated by the first antenna 342 and the second antenna 343 are the RTK corrected tillage depth change values.

[0087] Step S600: Using the trained tillage depth detection model, correct the tillage depth data of the real-time obtained zero-value benchmark, attitude correction value, indirect tillage depth value, and tillage depth change value; and

[0088] Step S700: The corrected tillage depth data is fused to obtain accurate online detection values ​​of paddy field tillage depth.

[0089] In step S700, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the Kalman fusion algorithm is used to fuse the filtered tillage depth data; or in step S600, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the trained tillage depth detection model is used to correct the tillage depth data.

[0090] In this embodiment, the process of establishing the tillage depth detection model is as follows. To simplify the problems encountered in actual operations, some assumptions are made to simplify the actual situation before establishing the model:

[0091] The established model does not consider the influence of mechanical clearance; the pitch attitude change of the tractor 1 during operation is centered on the contact point between the rear traveling wheel 12 of the tractor 1 and the ground; the roll attitude change of the tractor 1 during operation is centered on the installation position of the fuselage gyroscope 32.

[0092] See Figure 6 , Figure 6This is a schematic diagram of the pitch direction movement of the tractor 1 or rotary tiller 2 according to an embodiment of the present invention. In the pitch direction of the tractor 1, there are two rotational movements: one is the rotation around the contact point between the rear travel wheel 12 and the ground, detected by the gyroscope 32 mounted on the tractor 1, and the angle between the vehicle body and the horizontal plane is recorded as α; the other is the rotation around the hinge point of the lower cantilever of the three-point suspension, detected by the tilt sensor 33 mounted on the lower cantilever of the three-point suspension, and the rotation angle of the three-point suspension is recorded as θ. Therefore, when the tractor 1 body has no change in pitch angle, its tillage depth is:

[0093] h = l sin(θ).

[0094] See Figure 7 , Figure 7 This is a schematic diagram of the movement of the tractor 1 or rotary tiller 2 in the roll direction according to an embodiment of the present invention. Analysis of tillage depth during roll attitude changes shows that changes in the pitch direction have no effect on tillage depth during roll attitude. Based on the aforementioned assumption, the roll angle of the tractor 1 is the same as the change in roll angle of the rotary tiller 2. The roll angle detected by the gyroscope 32 is β. The schematic diagram of the vehicle body's movement in the roll direction is shown below. Figure 7 As shown. Meanwhile, the vehicle's heading angle, γ, can be detected by the BDS dual antennas.

[0095] The sinking analysis of the traction machine 1 is based on the difference in distance between the lowest point of the rear travel wheel 12 on one side of the traction machine 1 and the corresponding position point of the ground contouring mechanism 31. The undulation changes of the traction machine 1 body are obtained through the first antenna 342 on the top of the traction machine 1.

[0096] See Figure 8 , Figure 8 This is a schematic diagram illustrating the establishment of the coordinate system for a detection model according to an embodiment of the present invention. To express the model's computations, the above analysis is presented in coordinate system form. A geodetic coordinate system is established with the RTK base station 341 set up in the field as the origin, and the x, y, and z axes representing east, north, and the vertical upward direction, respectively. The coordinates are distinguished by adding a subscript D. A traction machine 1 coordinate system T is established with the fuselage gyroscope 32 on the traction machine 1 as the origin, and the x, y, and z axes representing the vehicle's forward direction, the direction to the right of the vehicle's forward direction, and the vertical upward direction, respectively. The coordinates are distinguished by adding a subscript T. The origins of both coordinate systems can be located using the BDS system. Their relative distances in the geodetic coordinate system (x, y, z) are respectively x... tf y tf , z tf .

[0097] In the coordinate system of traction machine 1, based on the mapping of traction machine 1, the coordinates of the lowest point of the rotary tiller blade at the corresponding position on the two contour gyroscopes 319 of the ground contouring mechanism 31 of rotary tiller 2 in the coordinate system of traction machine 1 can be obtained respectively:

[0098] [x Tld y Tld z Tld ] T , [x Trd y Trd z Trd ] T ;

[0099] The coordinates of the two contouring gyroscopes 319 of the ground contouring mechanism 31 and the ground contact point are as follows:

[0100] [x Tlf y Tlf z Tlf ] T , [x Trf y Trf z Trf ] T ;

[0101] According to the coordinate transformation formula, all points in the tractor 1 coordinate system can be transformed to the geodetic coordinate system. The transformation formula is as follows:

[0102]

[0103] Where s represents sin and c represents cos.

[0104] During the operation, the tillage depth has two measured values ​​at the same sampling time, which can be expressed as:

[0105]

[0106] The real-time tillage depth can be calculated using the above formula.

[0107] During field operations, the first antenna 342 of the BDS dual-antenna system 34 is mounted on the central axis of the top of the tractor 1 using a strong magnetic base and strong double-sided adhesive; the second antenna 343 is mounted on the central axis of the rotary tiller 2 using a strong magnetic base and strong double-sided adhesive; the body gyroscope 32 is mounted behind the cab of the tractor 1 using strong double-sided adhesive and cable ties; the tilt sensor 33 is mounted on the lower cantilever of the three-point suspension of the tractor 1 using strong double-sided adhesive and cable ties; the ground contouring mechanism 31 is fixedly connected to the tractor 1 with bolts through the body connector 311, and its matching contouring gyroscope 319 is fixedly connected to the upper crossbar 315 of the ground contouring mechanism 31 with bolts.

[0108] In this embodiment, before installing each sensor, a backpropagation (BP) network needs to be trained and constructed for the characteristics of each sensor, which is used for filtering the sensor data. The following explanation uses the relative position output of a BDS dual-antenna system (34-way BP network) as an example:

[0109] First, the true values ​​between the first antenna 342 and the second antenna 343 are measured. Then, the data is left to stand still and the output data is collected. The set true value is then changed, and data collection continues. The sensor sampling frequency is 5Hz, and the duration of each experiment is set, such as 5 minutes. After multiple sets (e.g., ten sets) of experiments, the above data is used as the dataset, and training and validation sets are established in a 6:1 ratio. During training, to avoid the influence of the raw data on network training, normalization is performed. The processed sensor data is used as the input value, and the measured true value is used as the label for BP network training. This completes the training of the sensor characteristic network.

[0110] After the installation and fixing of each sensor are completed, the ground contouring mechanism 31 and the tilt sensor 33 are calibrated. The calibration of the ground contouring mechanism 31 is performed on a level surface. The two support plates 317 of the ground contouring mechanism 31 are placed naturally on the ground. The data processor 35 is turned on, and data is collected for a set time, such as 2 minutes. The average value of the output data from the two contouring gyroscopes 319 is taken as the reference value for the level surface, completing the calibration of the ground contouring mechanism 31. The calibration of the tilt sensor 33 is performed on a level road surface. After the system is working normally, the traction machine 1 is run under three-point suspension lifting conditions of 20cm and 35cm and lowering conditions of 10cm and 20cm. Data is collected for 2 minutes under each condition, and the average value is taken as the effective value for each calibration. Then, using the angle of lifting and lowering of the three-point suspension detected by the tilt sensor 33 as x, and the actual height of lifting and lowering of the three-point suspension as y, where the lifting value is positive and the lowering value is negative, the corresponding relationship is obtained using the least squares method, and the calibration is completed. The lifting and lowering distances and the data collection time can be adjusted as needed. This is only an example and is not intended to limit the invention.

[0111] Before commencing work, an RTK base station 341 needs to be set up in the non-operational area outside the field to ensure that the positioning data calculated by the first antenna 342 and the second antenna 343 on the machine is RTK corrected data. Before the tractor 1 enters the field to begin work, the power to the data processor 35 is turned on to ensure that the data processor 35 can receive and store data. After the entire paddy field tillage depth detection device 3 is working properly, the tractor 1 begins rotary tillage. During the rotary tillage operation of the tractor 1, the data collected by each sensor is processed by the data processor 35 and finally stored on the SD card, providing a data foundation for subsequent planting, management, and harvesting.

[0112] See Figure 9 , Figure 9This is a schematic diagram of the BP-Kalman filtering algorithm according to an embodiment of the present invention. In this embodiment, the data collected by each sensor is first filtered using the BP-Kalman filtering algorithm. The BP-Kalman filtering algorithm is based on the Kalman algorithm, introducing a BP neural network to model it. The BP neural network mimics the neural system structure of the human brain, possessing self-learning and nonlinear mapping capabilities, and reduces errors through gradient descent. In this algorithm, a three-layer network structure is used for modeling with the BP neural network: an input layer, a hidden layer, and an output layer, with fully connected layers. Taking the BDS elevation difference data obtained in the calibration experiment as an example, the other sensors are processed in the same way. The acquired data are used to establish training and validation sets in a 6:1 ratio. During training, normalization is performed to avoid the influence of the original data on network training. The processed sensor data is used as input, and the experimentally recorded true values ​​are used as labels. This embodiment selects the SIGMOID function, which has nonlinear relationship characteristics, as the activation function.

[0113]

[0114] After training is complete, taking the relative distance output of the BDS dual-antenna system 34 as an example, the process involves using the initial input state vector and the error covariance matrix... P0, and then during the loop iteration, each time the previous prediction value is used. (Initial values ​​input in the first iteration) are input into the model to obtain the predicted values ​​for this iteration. Then, the transition matrix is ​​calculated using the following formula:

[0115]

[0116] Next, calculate the prior error covariance:

[0117]

[0118] In the formula P k-1 Let be the estimated and true values ​​of the error covariance matrix at time k and (k-1) respectively, and Q be the state noise covariance matrix;

[0119] Next, calculate the Kalman gain:

[0120]

[0121] In the formula K k Here, H represents the Kalman gain at time k, and H is the state output matrix.

[0122] Update the state estimate using the measured values:

[0123]

[0124] In the formula zk-1 The measurement vector at time (k-1);

[0125] Last updated error covariance:

[0126]

[0127] By repeating the above steps and continuously outputting the optimal value, the filtering of real-time data collected by each sensor can be completed.

[0128] In the above formulas, α is the angle between the fuselage of tractor 1 and the horizontal plane; θ is the rotation angle of the three-point suspension of tractor 1; h is the tillage depth of tractor 1 when there is no change in pitch angle; β is the roll angle of tractor 1; γ is the heading angle of tractor 1; [x Tld y Tld z Trd ] T , [x Trd y Trd z Trd ] T Let the lowest point on both sides of the rotary tiller be located in the coordinate system of the tractor; [x Tlf y Tlf z Tlf ] T , [x Trf y Trf z Trf ] T The coordinates of the two detection units of the contouring mechanism and the ground contact point; x tf y tf , z tf H represents the distance between the origins of the two coordinate systems in the x, y, z coordinates of the geodetic coordinate system; H is the corrected tillage depth value. This is the predicted value from the previous iteration; Here is the predicted value for this iteration; A is the transition matrix; P is the estimated value of the error covariance matrix at time k; k-1 Let K be the true value of the error covariance matrix at time (k-1); Q is the state noise covariance matrix; K k Z is the Kalman gain at time k; H is the state output matrix; z k-1 Let be the measurement vector at time (k-1).

[0129] Subsequently, the tillage depth values ​​output by the tilt sensor 33 and the BDS dual-antenna system 34 are fused using a Kalman fusion algorithm. The algorithm steps and principles are similar to the Kalman filter algorithm improved by the BP neural network. The calculated predicted value is replaced with the output value of another sensor, and then the Kalman gain is calculated to obtain the different proportions of the two sensors in the fusion process, and finally the data fusion is completed.

[0130] This invention uses multi-source sensor fusion technology to detect the tillage depth of rotary tillage in paddy fields. It employs a ground contouring mechanism 31, establishes a sensor layout and tillage depth detection model, and combines data filtering and fusion algorithms. The detected tillage depth data is corrected using the machine's sinking depth and attitude values, and the corrected data is then fused to obtain an accurate online detection value of paddy field tillage depth, thus achieving accurate online detection of paddy field tillage depth.

[0131] Of course, the present invention may have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and modifications according to the present invention, but these corresponding changes and modifications should all fall within the protection scope of the appended claims.

Claims

1. A paddy field tillage depth detection device, characterized in that, include: Ground contouring mechanism is used to detect changes in the depth of mud foot sinking and terrain elevation during tractor operation; An inclination sensor is used to detect changes in the inclination angle of the three-point suspension of the tractor to indirectly obtain tillage depth data; The fuselage gyroscope is used to detect the roll and pitch attitude of the tractor, so as to correct the tillage depth detection error caused by the attitude change of the tractor during operation. The BDS dual-antenna system is used to detect the absolute value of the working trajectory and tillage depth changes of the tractor. as well as The data processor is communicatively connected to the contouring mechanism, the BDS dual-antenna system, the fuselage gyroscope, and the tilt sensor, respectively, and receives and saves the data collected by the contouring mechanism, the BDS dual-antenna system, the fuselage gyroscope, and the tilt sensor; The ground contouring mechanism is a symmetrical structure, comprising: Horizontal support; The trays are symmetrically arranged on both sides of the center line of the horizontal support and are connected to the horizontal support through the connecting frame. The trays are installed at the bottom of the connecting frame and the bottom of the trays are in contact with the ground. A limiting plate is installed on the horizontal support and is connected and cooperated with the connecting frame to realize the degree of freedom of the pallet in the horizontal rolling direction; A contour-following gyroscope is mounted on the connecting frame; and A return spring is mounted on the connecting frame.

2. The paddy field tillage depth detection device as described in claim 1, characterized in that, The data processor uses a microcontroller development board and is connected to the contouring mechanism, BDS dual-antenna system, fuselage gyroscope and tilt sensor via CAN bus. The SDIO controller is used to save the collected data to an SD card.

3. The paddy field tillage depth detection device as described in claim 2, characterized in that, The data processor drives the LCD screen to display its working status in real time via FSMC.

4. The paddy field tillage depth detection device as described in claim 1, characterized in that, The connecting frame is a rectangular frame. The front vertical rod of the connecting frame is connected to the horizontal support, and the bottom end of the rear vertical rod of the connecting frame is hinged to the support plate to ensure that the support plate is in contact with the ground during the contouring process.

5. The paddy field tillage depth detection device as described in claim 4, characterized in that, The contouring gyroscope is mounted on the upper crossbar of the connecting frame and connected to the data processor. It is used to detect the angle change of the upper crossbar and obtain quantitative data on the sinking of the machine and the undulation of the terrain during the contouring process.

6. The paddy field tillage depth detection device as described in claim 4, characterized in that, The limiting plate is an arc-shaped plate, and both ends of the arc-shaped plate are connected to the horizontal support; the middle of the arc-shaped plate is an arc-shaped through hole, and the front vertical rod is provided with a limiting boss that matches the middle arc-shaped through hole.

7. A method for detecting the tillage depth of paddy fields, characterized in that, The paddy field tillage depth detection device according to any one of claims 1-6 is used, and includes the following steps: S100. The ground contouring mechanism is installed behind the tractor to detect the ground shape and the subsidence of the implements before tillage, which serves as the zero reference value for tillage depth. S200. Install a fuselage gyroscope directly behind the cab of the tractor to detect attitude changes during tractor operation as attitude correction values. S300. Install the tilt sensor on the three-point suspension arm of the traction machine to detect the tilt angle change of the three-point suspension arm and obtain the indirect tillage depth value. S400: The first and second antennas of the BDS dual-antenna system are respectively installed on the central axis of the tractor and rotary tiller roof to detect the change in tillage depth during the rotary tillage operation of the tractor. S500. Set up an RTK base station of the BDS dual-antenna system in a non-working area and connect it to the first antenna and the second antenna respectively, so that the data calculated by the first antenna and the second antenna are the RTK corrected tillage depth change values. S600, using the trained tillage depth detection model to correct the real-time acquired tillage depth zero-value benchmark, attitude correction value, indirect tillage depth value, and tillage depth change value; and S700 integrates the corrected tillage depth data to obtain accurate online detection values ​​for paddy field tillage depth.

8. The method for detecting paddy field tillage depth as described in claim 7, characterized in that, In step S700, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the Kalman fusion algorithm is used to fuse the filtered tillage depth data; or in step S600, the BP-Kalman filter algorithm is used to filter the corrected tillage depth data, and then the trained tillage depth detection model is used to correct the tillage depth data.

9. A paddy field tillage machine, characterized in that, The paddy field tillage depth detection device includes any one of claims 1-6.

Images