Non-contact monitoring system for stubble height of ratooning rice, control method and harvester
Through the combination of laser sensor, body inclination sensor and displacement sensor, the heading height is predicted by the support vector machine, and the problem of unstable pile height in the rice combined harvester is solved, and non-contact automatic control of the pile height of regenerated rice is realized , improves measurement accuracy and reliability and reduces labor costs.
Patent Information
- Application Number
- CN202411206579.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2044-08-30
AI Technical Summary
The existing rice combine harvesters lack automatic control function for cutting table height in recycled rice, resulting in unstable pile height, affecting the yield of recycled rice, and the sensor is easily disturbed, cost is high, and the electromagnetic reversing valve is easily damaged.
The laser sensor, body inclination sensor and displacement sensor are combined to process data through multi-stage threshold filtering and Kalman filtering, and the header height is predicted by a support vector machine, and the header height is automatically adjusted through a hydraulic cylinder controller.
It improves measurement accuracy, reduces interference from external factors, realizes non-contact automatic control of the pile height of recycled rice, saves labor costs, and improves data accuracy and reliability.
Smart Images

Figure CN119065304A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of automatic regulation of regenerated rice stubble height, and in particular relates to a non-contact monitoring system and control method for the regenerated rice stubble height, and a harvester. Background Art
[0002] Ratoon rice is a double-harvest rice, which refers to certain cultivation and management measures taken after the first season of rice harvest to promote the germination of dormant axillary buds on the piles, and finally heading, flowering and fruiting. In order to ensure the yield of the second season of ratoon rice, manual harvesting is often used to effectively increase the multiple cropping index during the first season. However, with the rapid development of my country's economy and society and the transfer of young and middle-aged labor in rural areas to cities, the use of rice combine harvesters instead of manual harvesting has become the main way to harvest ratoon rice in the first season. However, domestic rice combine harvesters generally do not have the function of automatic control of the height of the cutting table, and manual operation to adjust the cutting table to ensure the stability of the height of the pile is not only labor-intensive but also has problems such as unstable height of the pile, which potentially affects the overall yield of ratoon rice, resulting in damage to the axillary buds of the ratoon rice piles, slow development, and stunted growth, resulting in the loss of the second season of ratoon rice yield, which restricts the expansion of the scale of ratoon rice planting in my country. In addition, the pile height of the common ratoon rice varieties in my country currently varies from 250mm to 400mm due to variety and regional differences. Therefore, different heights of the header pile heights are required for different regions and varieties of regenerated rice. At present, the adjustment of the header height of the rice-wheat combine harvester during harvesting operation basically relies on the operator to operate the hydraulic control valve in real time according to the field conditions. Due to the complex and changeable working environment and conditions, the operator's experience and technical requirements are high. Therefore, designing an automatic detection and adjustment method for the height of the regenerated rice pile to protect the dominant dormant buds of regenerated rice and promote the increase of production and income in the regeneration season is still an important issue in the research and use of regenerated rice combine harvesters.
[0003] Most foreign countries use the header profiling mechanism to achieve automatic adjustment of the header height of the harvester, such as the contact header height profiling mechanism of the John Deere JD-1075H combine harvester, which obtains the height change information from the ground by swinging the profiling plate installed under the header with the ground undulation. Yang Yinhui designed a header height control system based on ultrasonic sensors, which detects the height by installing multiple ultrasonic sensors at the bottom of the header and uses an electro-hydraulic proportional valve for control. Liao Yong et al. used infrared sensors to measure the crop height, installed displacement sensors on the lifting cylinder to indirectly measure the header height, and used electromagnetic reversing valves to control the header lifting. However, there are still some problems in the above research, such as the method of installing a profiling plate under the header is greatly affected by terrain changes; the ultrasonic sensor is easily disturbed by crops, and the electro-hydraulic proportional valve used is relatively expensive; the method of measuring the header height by using the displacement of the oil cylinder is not well adapted to the ground undulation, and the control of the header lifting by the electromagnetic reversing valve is prone to overshoot, and the frequent switching of the electromagnetic valve is prone to damage to the valve core. In addition, the regenerated rice stumps are the stem residues left after the rice is harvested, mainly including the rice stems, leaf sheaths, leaves, etc. In the practice of mechanized harvesting of the first season of regenerated rice, it was found that due to the high height and large number of leaves of the regenerated rice stumps left after the first season of harvesting, it would cause certain interference to the sensor detection height, resulting in part of the received height data being the distance from the sensor installation position to the rice stems and leaves. Therefore, the received data needs to be processed to meet the needs of height monitoring.
[0004] Therefore, there is an urgent need for a low-cost and practical technical means that can improve measurement accuracy and reduce interference from external factors. Summary of the invention
[0005] In response to the above technical problems, the present invention provides a non-contact monitoring system and control method for the height of regenerated rice stubbles and a harvester, which improve measurement accuracy and reduce interference from external factors. The height of the harvesting platform can be automatically adjusted without human intervention, thereby controlling the height of the regenerated rice stubbles and saving labor costs.
[0006] The present invention achieves the above technical objectives through the following technical means.
[0007] A non-contact monitoring system for the height of regenerated rice piles, comprising a controller, a hydraulic cylinder controller, a vehicle body inclination sensor, a laser sensor and a displacement sensor;
[0008] The body inclination sensor is arranged on the harvester body and on the same side as the harvesting platform, and is used to collect the inclination angle information of the harvester body and transmit it to the controller;
[0009] The laser sensor is arranged on one side of the cutting platform and is used to collect the straight-line distance between the cutting platform and the ground and transmit it to the controller;
[0010] The displacement sensor is arranged on one side of the header height adjustment cylinder, and is used to collect the extension information of the header height adjustment cylinder and transmit it to the controller;
[0011] The controller is respectively connected to the hydraulic cylinder controller, the body inclination sensor, the laser sensor and the displacement sensor; the controller processes the inclination angle information of the harvester body, the straight-line distance of the cutting platform from the ground and the extension amount information of the cutting platform height adjustment cylinder, and controls the extension amount of the cutting platform height adjustment cylinder through the hydraulic cylinder controller to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0012] In the above scheme, the controller filters the data collected by the laser sensor and the displacement sensor through a combination of multi-stage threshold filtering and Kalman filtering; the controller performs real-time data correction on the filtered data and the data collected by the vehicle body inclination sensor to obtain data corrected based on multi-source sensors; the controller fuses the corrected data through a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the cutting platform through the height prediction model of the support vector machine; the controller adjusts the height of the cutting platform by controlling the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller according to the obtained predicted height value of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0013] A control method according to the non-contact monitoring system for the height of regenerated rice piles comprises the following steps:
[0014] Step S1, sensor data collection: the body inclination sensor collects the harvester body inclination angle information and transmits it to the controller, the laser sensor collects the straight-line distance of the cutting platform from the ground and transmits it to the controller, and the displacement sensor collects the extension amount information of the cutting platform height adjustment cylinder and transmits it to the controller;
[0015] Step S2, data filtering processing: the controller performs filtering processing on the data collected by the laser sensor and the displacement sensor in step S1 by combining multi-stage threshold filtering and Kalman filtering;
[0016] Step S3, data correction: the controller performs real-time data correction on the data obtained after filtering in step S2 and the data collected by the vehicle body inclination sensor to obtain data corrected based on multi-source sensors;
[0017] Step S4, data fusion: the controller fuses the corrected data obtained in step S3 by using a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the header by using a height prediction model of the support vector machine;
[0018] Step S5, control of the height of the cutting platform: the controller controls the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller according to the predicted height value of the cutting platform obtained in step S4 to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0019] In the above scheme, in the step S1 sensor data collection, when the change in the inclination angle of the harvester body collected by the body inclination sensor is less than 0.75°, the data is invalid; when the change in the inclination angle of the harvester body is ≥0.75°, the data is valid, and the inclination data is sent to the controller after the data collection is completed.
[0020] In the above scheme, the data filtering process in step S2 specifically includes the following steps:
[0021] Step S2.1, fitting the function between the extension of the header height adjustment cylinder output by the displacement sensor and the header height;
[0022] Step S2.2, first adopt multi-stage threshold filtering, based on the maximum and minimum values of the extension of the header height adjustment cylinder measured by the displacement sensor in the previous stage, set thresholds to process the height data measured by the laser sensor respectively, and limit the height measured by the laser sensor to the set threshold range;
[0023] Step S2.3: Use the Kalman filter method to perform Kalman filter processing on the height data after the multi-segment threshold filtering.
[0024] In the above scheme, the data correction in step S3 specifically includes the following steps:
[0025] First, calibrate the body inclination sensor on a horizontal plane and set it to zero. At this time, the data changes of the Z-axis and Y-axis of the body inclination sensor can respectively reflect the degree of inclination of the body in the front and rear and left and right directions. Set the angle change of the Y-axis to α and the change of the Z-axis to β. When the values change, the laser sensor will correct according to the increment, and the height data obtained is:
[0026] h z =h·cosα·cosβ
[0027] Where: h z is the height data of the laser sensor after correction, that is, the output value; h is the original height data collected by the laser sensor, that is, the input value.
[0028] In the above scheme, in step S4, the corrected data obtained in step S3 are fused by using the historical data adaptive weighting method:
[0029] The calculation formula of the height value at a certain moment after weighted average fusion is as follows:
[0030]
[0031]
[0032] Where: is the height value after filtering and weighted averaging at a certain moment, is the height value at the current moment after filtering, is the weight of a certain measurement data at the current moment, The mean absolute error of a set of measured data at the current moment; n is the number of data obtained; It is the current measurement value of a group of measurement data of the laser sensor; is the set height value, which is calculated based on the detection value of the displacement sensor; i is a certain moment.
[0033] Furthermore, the input values of the height prediction model in step S4 also include features related to height: average height and height change rate of regenerated rice within a preset number of days;
[0034] The height prediction model is:
[0035] y=w T φ(x)+b
[0036] in
[0037]
[0038] In the formula, x is the input vector, which contains the feature variables and historical data values, t represents the number of days, and h t is the average height on day t, r t is the rate of change of height between the tth day and the previous day, is the height value after filtering and weighted averaging at a certain moment, w is the weight vector obtained by model training, b is the bias term obtained by model training, φ(x) is the feature mapping function, and the RBF radial basis kernel function is used here, where x m is the support vector and σ is a hyperparameter related to the samples in the dataset.
[0039] In the above scheme, in step S5, after the controller adjusts the predicted height value of the cutting platform through the fuzzy PID controller, the height of the cutting platform is adjusted by controlling the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller, thereby adjusting the height of the regenerated rice pile.
[0040] A harvester comprises the non-contact monitoring system for the height of regenerated rice piles and is controlled according to the control method of the non-contact monitoring system for the height of regenerated rice piles.
[0041] Compared with the prior art, the present invention has the following beneficial effects:
[0042] The present invention uses a laser sensor, an inclination sensor, and a displacement sensor to collect information about the surrounding environment, identifies the straight-line distance of the header from the ground, the inclination of the vehicle body, and the displacement of the oil cylinder. The controller performs filtering processing on the data collected by the laser sensor and the displacement sensor by combining multi-segment threshold filtering and Kalman filtering; the controller performs real-time data correction on the data obtained after filtering and the data collected by the vehicle body inclination sensor to obtain data corrected based on multi-source sensors; the controller fuses the corrected data by a historical data adaptive weighting method, inputs it into a support vector machine, and predicts the predicted height value of the header through the height prediction model of the support vector machine; the controller controls the elongation of the header height adjustment cylinder through a hydraulic cylinder controller according to the obtained predicted height value of the header to adjust the height of the header, thereby adjusting the height of the regenerated rice pile. The present invention integrates multi-source sensing, realizes non-contact automatic control of the height of the regenerated rice pile, improves the accuracy and reliability of the data, and can automatically adjust the height of the header without manual participation, greatly saving labor costs and experimental accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 It is a schematic diagram of the assembly of a sensor according to one embodiment of the present invention on a harvester.
[0044] Figure 2 It is a schematic diagram of a method flow chart of an embodiment of the present invention.
[0045] Figure 3 It is a schematic diagram of the relationship between the installation plane of the tilt sensor and the XYZ axes according to one embodiment of the present invention.
[0046] In the figure: 1. Controller; 2. Hydraulic cylinder control system; 3. Vehicle body inclination sensor; 4. Laser sensor; 5. Displacement sensor. DETAILED DESCRIPTION
[0047] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.
[0048] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "front", "back", "left", "right", "up", "down", "axial", "radial", "vertical", "horizontal", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.
[0049] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0050] Figure 1 The figure shows a preferred embodiment of the non-contact control system for the height of the regenerated rice piles, which comprises a controller 1, a hydraulic cylinder controller 2, a vehicle body inclination sensor 3, a laser sensor 4 and a displacement sensor 5;
[0051] The body inclination sensor 3 is arranged on the harvester body and on the same side as the harvesting platform, and is used to collect the inclination angle information of the harvester body and transmit it to the controller 1;
[0052] The laser sensor 4 is arranged on the side plate of the cutting platform divider, and is used to collect the straight-line distance between the cutting platform and the ground and transmit it to the controller 1;
[0053] The displacement sensor 5 is arranged on one side of the header height adjustment cylinder, and is used to collect the extension information of the header height adjustment cylinder and transmit it to the controller 1;
[0054] The controller 1 is respectively connected to the hydraulic cylinder controller 2, the body inclination sensor 3, the laser sensor 4 and the displacement sensor 5; the controller 1 processes the inclination angle information of the harvester body, the straight-line distance of the cutting platform from the ground and the extension amount information of the cutting platform height adjustment cylinder, and controls the extension amount of the cutting platform height adjustment cylinder through the hydraulic cylinder controller 2 to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0055] The controller 1 performs filtering processing on the data collected by the laser sensor 4 and the displacement sensor 5 by combining multi-stage threshold filtering and Kalman filtering; the controller 1 performs real-time data correction on the filtered data and the data collected by the vehicle body inclination sensor 3 to obtain data corrected based on multi-source sensors; the controller 1 fuses the corrected data by a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the cutting platform by predicting the height prediction model of the support vector machine; the controller 1 adjusts the height of the cutting platform by controlling the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller 2 according to the obtained predicted height value of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0056] Preferably, the controller 1 is a single chip microcomputer.
[0057] like Figure 2 As shown, a control method according to the non-contact monitoring system for the height of regenerated rice piles comprises the following steps:
[0058] Step S1, sensor data collection: the body inclination sensor 3 collects the harvester body inclination angle information and transmits it to the controller 1, the laser sensor 4 collects the straight-line distance of the cutting platform from the ground and transmits it to the controller 1, and the displacement sensor 5 collects the extension amount information of the cutting platform height adjustment cylinder and transmits it to the controller 1;
[0059] Step S2, data filtering processing: the controller 1 performs filtering processing on the data collected by the laser sensor 4 and the displacement sensor 5 in step S1 by combining multi-stage threshold filtering and Kalman filtering;
[0060] Step S3, data correction: the controller 1 performs real-time data correction on the data obtained after filtering in step S2 and the data collected by the vehicle body inclination sensor 3 to obtain data corrected based on multi-source sensors;
[0061] Step S4, data fusion: the controller 1 fuses the corrected data obtained in step S3 by using a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the header by using a height prediction model of the support vector machine;
[0062] Step S5, control of the height of the cutting platform: the controller 1 controls the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller 2 according to the predicted height value of the cutting platform obtained in step S4 to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0063] The static measurement average value error of the inclination sensor 3 is ±0.01°. The inclination sensor 3 is a sensitive element. Since the harvester experiences slight bumps when driving in the field, in the sensor data collection step S1, when the change in the inclination angle of the harvester body collected by the inclination sensor 3 is less than 0.75°, the data is invalid; when the change in the inclination angle of the harvester body is ≥0.75°, the data is valid. After the data collection is completed, the inclination data is sent to the controller 1.
[0064] The data filtering process in step S2 specifically includes the following steps:
[0065] Step S2.1, fitting the function between the extension of the header height adjustment cylinder and the header height output by the displacement sensor 5 based on the least square method, and obtaining a more accurate result. When the cylinder is extended, the change of the header height can be obtained according to the function between the extension of the header height adjustment cylinder and the header height, and a more accurate result can be obtained;
[0066] Furthermore, in a specific embodiment of the present invention, in order to fit the function between the extension of the cutter height adjustment cylinder and the cutter height output by the displacement sensor 5, the cutter is manually raised from the lowest point to the highest point at intervals of 5 cm during calibration, and the cutter height and the voltage value of the displacement sensor are recorded during the process to obtain a linear function.
[0067] Step S2.2, first adopt multi-stage threshold filtering, adjust the maximum and minimum values of the extension of the oil cylinder based on the height of the header measured by the displacement sensor 5, set thresholds to process the height data measured by the laser sensor 4 respectively, and limit the height measured by the laser sensor 4 to the set threshold range;
[0068] Furthermore, based on the previous test conditions and the size of the pile height and the height of the blades from the ground, we first observed and analyzed them, and used multi-threshold filtering to set multiple thresholds, and divided the obtained height data into three data areas, which can effectively avoid the interference of field leaves on the detection data. Based on the maximum and minimum values of the extension of the header height adjustment cylinder measured by the displacement sensor 5, the thresholds are set to process the height data measured by the laser sensor 4 respectively, and the height measured by the laser sensor 4 is limited to the set threshold range; in particular, the upper and lower thresholds will change with the maximum and minimum values of the extension of the header height adjustment cylinder measured by the displacement sensor 5. The upper and lower thresholds will change with the regional segmentation as follows:
[0069]
[0070] In the formula, h i is the data measured by the laser sensor, H1 is the minimum value measured by the displacement sensor, and H2 is the maximum value measured by the displacement sensor.
[0071] Step S2.3: Use the Kalman filtering method to perform Kalman filtering on the height data after the multi-segment threshold filtering, and input the filtered data as sample data to the controller 1.
[0072] The present invention firstly adopts multi-stage threshold filtering combined with Kalman filtering to effectively eliminate abnormal values and noise, especially when the altitude data is interfered by external factors.
[0073] The step S3 data correction specifically includes the following steps:
[0074] First, calibrate the vehicle body inclination sensor 3 on a horizontal plane and set it to zero. Figure 3 As shown, at this time, the data changes of the Z axis and Y axis of the vehicle body inclination sensor 3 can respectively reflect the degree of inclination of the vehicle body in the front and rear and left and right directions. The angle change of the Y axis is set to α and the change of the Z axis is set to β. When the values change, the laser sensor 4 will be corrected according to the increment, and the height data obtained is:
[0075] h z =h·cosα·cosβ
[0076] Where: h z is the height data of the laser sensor 4 after correction, that is, the output value, in cm; h is the original height data collected by the laser sensor 4, that is, the input value, in mm.
[0077] The data fusion of step S4 comprises the following steps:
[0078] Laser sensor 4 data fusion based on historical data: by calculating the average absolute error between the measurement data of the laser sensor 4 in the period before the measurement and the true value of the header height, and controlling the number of acquired data, combined with the current height value, the height value at a certain moment after weighted average fusion is obtained;
[0079] The calculation formula of the height value at a certain moment after weighted average fusion is as follows:
[0080]
[0081] Where: is the height value after filtering and weighted averaging at a certain moment, is the height value at the current moment after filtering, is the weight of a certain measurement data at the current moment, The mean absolute error of a set of measured data at the current moment; n is the number of acquired data, which is 20; is the current measurement value of a group of measurement data of the laser sensor 4; is the set height value, which is calculated based on the detection value of the displacement sensor 5; i is a certain moment.
[0082] The support vector machine in step S4 uses support vector regression (SVR) to predict the height, which is specifically the following steps:
[0083] 1. Data collection: collect altitude data, which is the altitude value and related features after filtering and weighted averaging at a certain moment;
[0084] 2. Data preprocessing: processing missing values, outliers (such as extreme values), and feature standardization or normalization of the collected data;
[0085] 3. Divide the data set: Divide the preprocessed data into a training set and a test set. Specifically, 80% is the training set and 20% is the test set.
[0086] 4. Model training: Use the support vector regression (SVR) algorithm to train the height prediction model based on the training set;
[0087] 5. Model prediction: Use the trained height prediction model to make predictions on the test set;
[0088] 6. Evaluate the model: Use mean square error (MSE), coefficient of determination (R 2 ) to evaluate the model performance.
[0089] The height-related features include the average height and height change rate of the regenerated rice within a preset number of days.
[0090] The height prediction model in step S4 is:
[0091] y=w T φ(x)+b
[0092] in
[0093]
[0094] In the formula, x is the input vector, which contains the feature variables and historical data values, t represents the number of days, and h t is the average height on day t, r t is the rate of change of height between the tth day and the previous day, is the height value after filtering and weighted averaging at a certain moment, w is the weight vector obtained by model training, b is the bias term obtained by model training, φ(x) is the feature mapping function, and the RBF radial basis kernel function is used here, where x m is the support vector and σ is a hyperparameter related to the samples in the dataset.
[0095] The adaptive weighted method for historical data of the present invention fuses the corrected data obtained in step S3, and predicts the predicted height value of the harvesting platform through the height prediction model of the support vector machine. The adaptive weighted method for historical data is combined with the support vector machine, which helps to achieve efficient, flexible and accurate prediction in the process of height data processing.
[0096] In step S5, after the controller 1 adjusts the predicted height value of the cutting platform through the fuzzy PID controller, the height of the cutting platform is adjusted by controlling the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller 2, thereby adjusting the height of the regenerated rice pile.
[0097] In step S5, the controller 1 adjusts the predicted height value of the header through the fuzzy PID controller, specifically:
[0098] First, define the input and output. Input: Height error (e): The displacement sensor 5 uses the measured data after function calculation as the actual value, and the external input predicted value r as the target value to obtain the height error e (e = cr) and the height error change rate ec (ec = de / dt). Output: Proportional coefficient K P , integral coefficient K I , differential coefficient K D : PID control signal corresponding to the adjustment effect.
[0099] Then, define the fuzzy set: the fuzzy subsets of the input and output variables are quantized into 7 levels, and their fuzzy word sets are represented as: [NBNM NS ZO PS PM PB]. The basic domain is {-6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6}.
[0100] Next, design fuzzy rules: According to the input combination, determine the control signal output in each case. The specific fuzzy rule table is as follows (cited from Liu Weijian, Luo Xiwen, Zeng Shan, et al. Performance test and analysis of adaptive profiling harvesting platform for regenerated rice based on fuzzy PID control [J]. Transactions of the Chinese Society of Agricultural Engineering, 2022, 38(10): 1-9.):
[0101] Table 1 Fuzzy rules table
[0102]
[0103] Finally, defuzzification: convert the fuzzy output into the proportional coefficient K of the PID controller parameters P , integral coefficient K I , differential coefficient K D As the output of the fuzzy PID controller.
[0104] Furthermore, the controller 1 performs PID processing on the data based on the result obtained by the fuzzy PID controller, and inputs the processed data into the hydraulic cylinder controller 2 to control the extension of the cutting platform height adjustment cylinder to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
[0105] The fuzzy PID algorithm of the fuzzy PID controller of the present invention is used for regulation and control, which can improve the overall performance and adaptability of the system.
[0106] A harvester comprises the non-contact monitoring system for the height of regenerated rice piles and is controlled according to the control method of the non-contact monitoring system for the height of regenerated rice piles.
[0107] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.
[0108] The series of detailed descriptions listed above are only specific descriptions of feasible embodiments of the present invention. They are not intended to limit the scope of protection of the present invention. All equivalent embodiments or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.
Claims
1. A non-contact monitoring system for the height of regenerated rice piles. It is characterized in that It comprises a controller (1), a hydraulic cylinder controller (2), a vehicle body tilt sensor (3), a laser sensor (4) and a displacement sensor (5); The body inclination sensor (3) is arranged on the harvester body and on the same side as the harvesting platform, and is used to collect the inclination angle information of the harvester body and transmit it to the controller (1); The laser sensor (4) is arranged on one side of the cutting platform and is used to collect the straight-line distance between the cutting platform and the ground and transmit it to the controller (1); The displacement sensor (5) is arranged on one side of the header height adjustment cylinder and is used to collect information on the extension of the header height adjustment cylinder and transmit it to the controller (1); The controller (1) is respectively connected to a hydraulic cylinder controller (2), a body inclination sensor (3), a laser sensor (4) and a displacement sensor (5); the controller (1) processes information on the inclination angle of the harvester body, the straight-line distance of the cutting platform from the ground and the extension amount of the cutting platform height adjustment cylinder, and controls the extension amount of the cutting platform height adjustment cylinder through the hydraulic cylinder controller (2) to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
2. The non-contact monitoring system for the height of regenerated rice piles according to claim 1, It is characterized in that The controller (1) performs filtering processing on the data collected by the laser sensor (4) and the displacement sensor (5) by combining multi-stage threshold filtering and Kalman filtering; the controller (1) performs real-time data correction on the filtered data and the data collected by the vehicle body tilt sensor (3) to obtain data corrected based on multi-source sensors; the controller (1) fuses the corrected data by using a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the cutting platform by using a height prediction model of the support vector machine; the controller (1) controls the extension of the cutting platform height adjustment cylinder through a hydraulic cylinder controller (2) according to the obtained predicted height value of the cutting platform to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
3. A control method for a non-contact monitoring system for regenerated rice pile height according to any one of claims 1 to 2, It is characterized in that The following steps are involved: Step S1, sensor data collection: the body inclination sensor (3) collects the harvester body inclination angle information and transmits it to the controller (1), the laser sensor (4) collects the straight-line distance of the cutting platform from the ground and transmits it to the controller (1), and the displacement sensor (5) collects the extension amount information of the cutting platform height adjustment cylinder and transmits it to the controller (1); Step S2, data filtering processing: the controller (1) performs filtering processing on the data collected by the laser sensor (4) and the displacement sensor (5) in step S1 by combining multi-stage threshold filtering and Kalman filtering; Step S3, data correction: the controller (1) performs real-time data correction on the data obtained after filtering in step S2 and the data collected by the vehicle body tilt sensor (3), to obtain data corrected based on multi-source sensors; Step S4, data fusion: the controller (1) fuses the corrected data obtained in step S3 by using a historical data adaptive weighting method, inputs the data into a support vector machine, and obtains a predicted height value of the header by using a height prediction model of the support vector machine; Step S5, control of the height of the cutting platform: the controller (1) controls the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller (2) according to the predicted height value of the cutting platform obtained in step S4 to adjust the height of the cutting platform, thereby adjusting the height of the regenerated rice pile.
4. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 3, It is characterized in that In the sensor data collection step S1, when the change in the inclination angle of the harvester body collected by the inclination angle sensor (3) is less than 0.75°, the data is invalid; when the change in the inclination angle of the harvester body is greater than or equal to 0.75°, the data is valid, and after the data collection is completed, the inclination angle data is sent to the controller (1).
5. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 3, It is characterized in that The data filtering process in step S2 specifically includes the following steps: Step S2.1, fitting the function between the extension of the header height adjustment cylinder output by the displacement sensor (5) and the header height; Step S2.2, first adopt multi-stage threshold filtering, based on the maximum and minimum values of the extension of the cutting platform adjustment cylinder measured by the displacement sensor (5), set thresholds to process the height data measured by the laser sensor (4), and limit the height measured by the laser sensor (4) to within the set threshold range; Step S2.3: Use the Kalman filter method to perform Kalman filter processing on the height data after the multi-segment threshold filtering.
6. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 3, It is characterized in that The step S3 data correction specifically includes the following steps: First, the vehicle body tilt angle sensor (3) is calibrated on a horizontal plane and set to zero. At this time, the data changes of the Z axis and Y axis of the vehicle body tilt angle sensor (3) can respectively reflect the tilt degree of the vehicle body in the front and rear and left and right directions. The angle change of the Y axis is set to α and the change of the Z axis is set to β. When the values change, the laser sensor (4) will be corrected according to the increment, and the height data obtained is: h z =h·cosα·cosβ Where: h z is the height data of the laser sensor (4) after correction, that is, the output value; and h is the original height data collected by the laser sensor (4), that is, the input value.
7. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 3, It is characterized in that In step S4, the corrected data obtained in step S3 are fused by using the historical data adaptive weighting method: The calculation formula of the height value at a certain moment after weighted average fusion is as follows: Where: is the height value after filtering and weighted averaging at a certain moment, is the height value at the current moment after filtering, is the weight of a certain measurement data at the current moment, The mean absolute error of a set of measured data at the current moment; n is the number of data obtained; is the current measurement value of a group of measurement data of the laser sensor (4); is the set height value, which is calculated based on the detection value of the displacement sensor (5); i is a certain moment.
8. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 7, It is characterized in that The input values of the height prediction model in step S4 also include features related to height: the average height and height change rate of regenerated rice within a preset number of days; The height prediction model is: y=w T φ(x)+b in In the formula, x is the input vector, which contains the feature variables and historical data values, t represents the number of days, and h t is the average height on day t, r t is the rate of change of height between the tth day and the previous day, is the height value after filtering and weighted averaging at a certain moment, w is the weight vector obtained by model training, b is the bias term obtained by model training, φ(x) is the feature mapping function, and the RBF radial basis kernel function is used here, where x m is the support vector and σ is a hyperparameter related to the samples in the dataset.
9. The control method of the non-contact monitoring system for the height of regenerated rice piles according to claim 7, It is characterized in that In step S5, after the controller (1) adjusts the predicted height value of the cutting platform through the fuzzy PID controller, the height of the cutting platform is adjusted by controlling the extension of the cutting platform height adjustment cylinder through the hydraulic cylinder controller (2), thereby adjusting the height of the regenerated rice pile.
10. A harvester, It is characterized in that It comprises the non-contact monitoring system for the height of regenerated rice piles according to claim 1 or 2, and is controlled according to the control method for the non-contact monitoring system for the height of regenerated rice piles according to any one of claims 3-9.
Citation Information
Patent Citations
Harvester header height automatic adjusting system and control method
CN106233921A
Automatic copying follow-up system of sugarcane harvester and control method thereof
CN110352707A
Profiling anti-blocking leaf vegetable harvester and control system and method control thereof
CN111226583A
Header self-adaptive adjusting rice harvester
CN112690089A
Electric power tower inclination monitoring and analyzing system based on Kalman filtering
CN117451006A
Cited By
Highland barley low-loss header real-time cutting depth adjusting method based on intelligent sensor
CN119885109A
Ratoon rice stubble height online detection and header dynamic regulation and control method and system and harvester
CN122074291A
