Real-time feed amount prediction method and system for combine harvester based on sensor fusion

Abstract

The present invention discloses a method and system for predicting the real-time feeding amount of a combine harvester based on sensor fusion. The method first establishes a first observation equation of the cross-bridge torque z1 and the feeding amount x through experiments, establishes a second observation equation of the traveling speed z2 of the combine harvester and the feeding amount x according to the definition of the feeding amount, and establishes a state equation of the torque of the combine harvester: x t =x t‑1 +Δx. Then, real-time detection values of the cross-bridge torque and the traveling speed are obtained; finally, the Kalman filtering algorithm is used to fuse the traveling speed and the cross-bridge torque to achieve the prediction of the feeding amount of the combine harvester. The present invention can achieve accurate measurement of the real-time feeding amount of the combine harvester.

Description

Technical Field

[0001] The present invention relates to the technical field of combine harvesters, and in particular to a method and system for predicting the real-time feed amount of a combine harvester based on sensor fusion. Background Art

[0002] Combine harvester feed volume refers to the total amount of crops that a combine harvester can process per unit time. Combine harvester feed volume is a very important design parameter that characterizes the operating capacity of a combine harvester, and is also an important basis for the design and calculation of each working component. The real-time feed volume of a combine harvester during operation is an important parameter that affects the quality and efficiency of grain harvesting operations. Too little feed volume will result in insufficient machine load and reduced operating efficiency; too much feed volume will overload the machine, and even cause blockage and failure. At the same time, too much or too little feed volume will also affect the operating quality of the combine harvester, resulting in excessively high impurity and loss rates of the combine harvester.

[0003] The calculation of the real-time feed amount of the combine harvester is shown in the following formula: F = ρLV Formula 1;

[0004] Where F is the real-time feed rate of the combine harvester, kg / s (kilograms per second); ρ is the total mass of grains and stalks per unit area in the uniform growth area at a certain cutting height, kg / m 2 (kg / m2); L is the width of the combine harvester’s cutting table, m (meter); V is the forward speed of the combine harvester, m / s (meter / second).

[0005] It can be seen from Formula 1 that the factors affecting the real-time feeding amount of the combine harvester include the operating speed of the combine harvester, the width of the combine harvester cutting table, and the crop planting density.

[0006] At present, in the actual production process, the feed amount of the combine harvester can be predicted by monitoring the torque of the combine harvester's cross-bridge driving shaft. However, there are the following problems in only measuring the torque of the combine harvester's cross-bridge driving shaft through a torque sensor and using it to predict the real-time feed amount of the combine harvester:

[0007] 1. The torque of the driving shaft entering the bridge reflects the amount of grain currently being processed in the bridge, which has a certain lag compared to the real-time feeding amount;

[0008] 2. Due to factors such as vibration, friction, and uneven material distribution, the output value of the cross-bridge torque sensor generally fluctuates greatly. If it is not processed, it cannot directly reflect the real-time feeding amount of the harvester.

[0009] 3. Due to factors such as terrain undulations, driver operation, and uneven material distribution, if the driving speed is used to measure the feed amount, the measurement will fluctuate greatly and cannot directly reflect the real-time feed amount.

[0010] Therefore, there is an urgent need for a method and system that can accurately measure the real-time feeding amount of a combine harvester. Summary of the invention

[0011] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a method for predicting the real-time feed amount of a combine harvester based on sensor fusion, which can achieve accurate measurement of the real-time feed amount of the combine harvester.

[0012] A method for real-time feed amount prediction of a combine harvester based on sensor fusion according to an embodiment of the present invention comprises the following steps:

[0013] S1. Through experiments, the first observation equation of the cross-bridge torque z1 and the feed amount x is established. According to the definition of the feed amount, the second observation equation of the combine harvester driving speed z2 and the feed amount x is established, and the state equation of the combine harvester torque x is established. t =x t-1 +Δx, where x t is the feed amount of the combine harvester at time t, Δx is the change in the feed amount of the combine harvester from time t-1 to time t;

[0014] S2. Acquire real-time detection values ​​of the bridge torque and the driving speed; wherein the bridge torque of the combine harvester can be collected by a torque sensor, which is installed on the bridge driving shaft of the combine harvester; the driving speed of the combine harvester can be collected by a speed monitoring module, which can be arranged in a measuring unit, and the measuring unit can be installed in the cab;

[0015] S3. The Kalman filter algorithm is used to fuse the detection values ​​of the driving speed and the bridge torque to realize the prediction of the feed amount of the combine harvester.

[0016] According to the method for predicting the real-time feed amount of a combine harvester based on sensor fusion according to an embodiment of the present invention, firstly, according to the relationship between the cross-bridge torque, the driving speed and the feed amount, a first observation equation, a second observation equation and a state equation are established; then, the real-time detection values ​​of the cross-bridge torque and the driving speed are obtained through the torque sensor and the speed monitoring module, and then the cross-bridge torque and the driving speed which are closely related to the feed amount of the harvester are data-fused through the Kalman filter algorithm to obtain the real-time value of the feed amount of the harvester. The data after Kalman filtering has a small jump and can accurately follow the changes in the feed amount, so that an accurate feed amount can be obtained.

[0017] In some embodiments of the present invention, the first observation equation is established by using an experimental calibration method in step S1, and the establishment method is as follows:

[0018] Step S11, obtaining the average planting density ρ of the field to be harvested, then the real-time feeding amount x of the combine harvester is: x=ρLV, where V is the travel speed of the combine harvester, and L is the width of the cutting table of the combine harvester;

[0019] Step S12, make the combine harvester travel a preset distance stably and evenly at a specified feed rate, and record the bridge torque values ​​at different feed rates, calculate the average value μ of the corresponding torque values, and perform linear regression on the data to obtain the first observation equation of the bridge torque z1 and the feed rate x: z1=H1(x)+B Formula 2;

[0020] Wherein, z1 is the measured value of the bridge torque, H1 is the slope of the linear regression, and B is the intercept of the linear regression; transforming Formula 2, we obtain z'1=z1-B=H1(x) Formula 3;

[0021] When the influence of observation noise V1 is added to the measurement, the value of the bridge torque z "1 after adding the observation noise V1 is obtained:

[0022] z "1=H1(x)+V1 Formula 4; where P(V1)=N(0,R1).

[0023] In an embodiment of the present invention, the method for obtaining the average planting density ρ of the field to be harvested in step S11 is:

[0024] In the field to be harvested, five 1m 3 The straw in the field is taken as the weight of grain, and the grain weight is m, then the average planting density of the field to be harvested is ρ=m / 5.

[0025] In an embodiment of the present invention, the method for establishing the second observation equation in step S1 is:

[0026] According to the real-time feed amount x=ρLV of the combine harvester and the influence of noise V2 in the measurement, the second observation equation of the travel speed z2 and the feed amount x is obtained: z2=H2(x)+V2 Formula 5;

[0027] Among them, P(V2)=N(0,R2).

[0028] In the embodiment of the present invention, the state equation of the torque of the combine harvester is established in step S1, x t =x t-1 +ρLΔV+W Formula 6; where P(W)=N(0,Q), W is the process noise, and ΔV is the speed change from time t-1 to time t.

[0029] In an embodiment of the present invention, the step S3 of using the Kalman filter algorithm to predict the feed amount of the combine harvester includes the following steps:

[0030] S31. Establish a priori estimate of the feed amount of the combine harvester at time K: Formula 7;

[0031] in, is the prior estimate of the combine harvester feed quantity at time K; is the posterior estimate of the combine harvester feed quantity at time K-1;

[0032] S32. Establish a priori estimate of the covariance at time K: Formula 8;

[0033] in, is the prior estimate of the state estimation covariance matrix at time K; is the a posteriori estimate of the state estimation covariance matrix at time K-1; Q is the variance of the process noise;

[0034] S33, update Kalman gain: Formula 9;

[0035] Among them, K K is the Kalman gain at time K; H is the observation matrix, \left [ {H1,H2} \right ] ; R is the covariance matrix of observation noise ;

[0036] S34, update the posterior estimation of the real-time feeding amount of the combine harvester at time K: Formula 10;

[0037] in, is the posterior estimation of the real-time feeding amount of the combine harvester at time K; k is the measured value of the feed amount at time K ;

[0038] S35. Update the posterior estimate P of the covariance at time K K : Formula 11.

[0039] The present invention also provides a system for real-time feed amount prediction of a combine harvester based on sensor fusion, comprising:

[0040] A torque sensor, wherein the torque sensor is installed on the cross-bridge driving shaft of the combine harvester and is used to measure the torque of the cross-bridge driving shaft of the combine harvester;

[0041] A measuring unit, the measuring unit being installed in a cab of the combine harvester and comprising a speed monitoring module for measuring a travel speed of the combine harvester;

[0042] A data processing unit, used to establish a state equation, a first observation equation and a second observation equation according to data from the torque sensor and the measurement unit, and perform data fusion using a Kalman filter algorithm;

[0043] The output unit is used to output the real-time feed amount prediction result of the combine harvester.

[0044] In an embodiment of the present invention, the data processing unit further includes:

[0045] Mean smoothing filter module, used for smoothing the output data of the torque sensor and speed monitoring module;

[0046] The Kalman filter module is used to integrate the driving speed and bridge torque data and output accurate feed amount prediction results.

[0047] In an embodiment of the present invention, it further comprises: a control unit, which is used to adjust the working parameters of the combine harvester according to the feed amount prediction result to achieve real-time control of the feed amount.

[0048] In an embodiment of the present invention, it also includes: a storage unit for storing data of the torque sensor, the speed monitoring module and the feed amount prediction result for subsequent analysis and optimization.

[0049] According to the sensor fusion-based real-time feed amount prediction system for a combine harvester according to an embodiment of the present invention, the bridge torque and driving speed of the combine harvester are detected in real time, and the state equation, the first observation equation and the second observation equation are established according to the relationship between the bridge torque, the driving speed and the feed amount, and the Kalman filter algorithm is used for data fusion to obtain the real-time value of the harvester feed amount. The data after Kalman filtering has a small jump and can accurately follow the changes in the feed amount. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a flow chart of a method for real-time feed amount prediction of a combine harvester based on sensor fusion according to an embodiment of the present invention;

[0051] Figure 2 is a schematic diagram of a system for real-time feed amount prediction of a combine harvester based on sensor fusion according to an embodiment of the present invention;

[0052] Figure 3 It is the relationship diagram between the torque sensor and the feed amount after smoothing and filtering;

[0053] Figure 4 This is a flow chart of the present invention using Kalman filter algorithm to calculate the feeding amount of a combine harvester;

[0054] Figure 5 This is a schematic diagram of the positions of the five-point sampling method. DETAILED DESCRIPTION

[0055] 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 are 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.

[0056] Reference below Figure 1-Figure 5 A method for real-time feed amount prediction of a combine harvester based on sensor fusion according to an embodiment of the present invention is described, comprising the following steps:

[0057] S1. Through experiments, the first observation equation of the cross-bridge torque z1 and the feed amount x is established. According to the definition of the feed amount, the second observation equation of the combine harvester driving speed z2 and the feed amount x is established, and the state equation of the combine harvester torque x is established. t =x t-1 +Δx, where x t is the feed amount of the combine harvester at time t, Δx is the change in the feed amount of the combine harvester from time t-1 to time t;

[0058] S2. Acquire real-time detection values ​​of the bridge torque and the driving speed; wherein the bridge torque of the combine harvester can be collected by a torque sensor, which is installed on the bridge driving shaft of the combine harvester; the driving speed of the combine harvester can be collected by a speed monitoring module, which can be arranged in a measuring unit, and the measuring unit can be installed in the cab;

[0059] S3. The Kalman filter algorithm is used to fuse the detection values ​​of the driving speed and the bridge torque to achieve real-time prediction of the feed amount of the combine harvester.

[0060] According to the method for predicting the real-time feed amount of a combine harvester based on sensor fusion according to an embodiment of the present invention, firstly, according to the relationship between the cross-bridge torque, the driving speed and the feed amount, a first observation equation, a second observation equation and a state equation are established; then, the real-time detection values ​​of the cross-bridge torque and the driving speed are obtained, and then, the cross-bridge torque and the driving speed which are closely related to the feed amount of the harvester are data-fused through the Kalman filter algorithm to obtain the real-time value of the feed amount of the harvester. The data after Kalman filtering has a small jump and can accurately follow the changes in the feed amount, so that an accurate feed amount can be obtained.

[0061] Considering the factors such as vibration, friction, and uneven material distribution, the output value of the cross-bridge torque sensor generally fluctuates greatly. If it is not processed, it cannot directly reflect the real-time feeding amount of the harvester. Therefore, a first-in-first-out queue can be constructed to perform sliding mean filtering to process the torque sensor data. In some embodiments of the present invention, in step S2, the first observation equation is established by using the experimental calibration method, and the output of the torque sensor is subjected to mean smoothing filtering when the first observation equation is established.

[0062] Specifically, the method for establishing the first observation equation is as follows:

[0063] Step S11, obtaining the average planting density ρ of the field to be harvested, then the real-time feeding amount x of the combine harvester is: x=ρLV, where V is the travel speed of the combine harvester, and L is the width of the cutting table of the combine harvester;

[0064] Step S12, make the combine harvester travel a preset distance stably and evenly at a specified feed rate, and record the bridge torque values ​​at different feed rates, calculate the average value μ of the corresponding torque values, and perform linear regression on the data to obtain the first observation equation of the bridge torque z1 and the feed rate x: z1=H1(x)+B Formula 2;

[0065] Wherein, z1 is the measured value of the bridge torque, H1 is the slope of the linear regression, and B is the intercept of the linear regression; transforming Formula 2, we obtain z'1=z1-B=H1(x) Formula 3;

[0066] When the influence of observation noise V1 is added to the measurement, the value of the bridge torque z "1 after adding the observation noise V1 is obtained:

[0067] z "1=H1(x)+V1 Formula 4; where P(V1)=N(0,R1).

[0068] The torque sensor data is processed by mean filtering, and the processing results are as follows Figure 3 shown.

[0069] In the embodiments of the present invention, referring to Figure 5 As shown, the method for obtaining the average planting density ρ of the field to be harvested in step S11 is:

[0070] In the field to be harvested, five 1m 3 The average planting density of the field to be harvested is ρ = m / 5. The five-point sampling method can be referred to Figure 5 The five points selected are: the center point of the field to be harvested, and the four midpoints between the center and the four corners.

[0071] In some embodiments of the present invention, the second observation equation is established by using the feed amount definition in step S1, and the establishment method is:

[0072] According to the real-time feed amount x=ρLV of the combine harvester and the influence of noise V2 in the measurement, the second observation equation of the travel speed z2 and the feed amount x is obtained: z2=H2(x)+V2 Formula 5;

[0073] Among them, P(V2)=N(0,R2).

[0074] In some embodiments of the present invention, the state equation of the torque of the combine harvester established in step S1 is: t =x t-1 +ρLΔV+W Formula 6; where P(W)=N(0,Q), W is the process noise, and ΔV is the speed change from time t-1 to time t.

[0075] In some embodiments of the present invention, the prediction of the feed amount of the combine harvester using the Kalman filter algorithm in step S3 includes the following steps:

[0076] S31. Establish a priori estimate of the feed amount of the combine harvester at time K: Formula 7; where is the prior estimate of the combine harvester feed quantity at time K; is the posterior estimate of the combine harvester feed quantity at time K-1;

[0077] S32. Establish a priori estimate of the covariance at time K: Formula 8;

[0078] in, is the prior estimate of the state estimation covariance matrix at time K; is the a posteriori estimate of the state estimation covariance matrix at time K-1; Q is the variance of the process noise;

[0079] S33, update Kalman gain: Formula 9;

[0080] Among them, K K is the Kalman gain at time K; H is the observation matrix, \left [ {H1,H2} \right ] ; R is the covariance matrix of observation noise ;

[0081] S34, update the posterior estimation of the real-time feeding amount of the combine harvester at time K: Formula 10;

[0082] in, is the posterior estimation of the real-time feeding amount of the combine harvester at time K; k is the measured value of the feed amount at time K ;

[0083] S35. Update the posterior estimate P of the covariance at time K K : Formula 11.

[0084] The present invention also provides a system for real-time feed amount prediction of a combine harvester based on sensor fusion, comprising:

[0085] A torque sensor, wherein the torque sensor is installed on the cross-bridge driving shaft of the combine harvester and is used to measure the torque of the cross-bridge driving shaft of the combine harvester;

[0086] A measuring unit, the measuring unit being installed in a cab of the combine harvester and comprising a speed monitoring module for measuring a travel speed of the combine harvester;

[0087] A data processing unit, used to establish a state equation, a first observation equation and a second observation equation according to data from the torque sensor and the measurement unit, and perform data fusion using a Kalman filter algorithm;

[0088] The output unit is used to output the real-time feed amount prediction result of the combine harvester.

[0089] In an embodiment of the present invention, the data processing unit further includes:

[0090] Mean smoothing filter module, used for smoothing the output data of the torque sensor and speed monitoring module;

[0091] The Kalman filter module is used to integrate the driving speed and bridge torque data and output accurate feed amount prediction results.

[0092] In an embodiment of the present invention, it further comprises: a control unit, which is used to adjust the working parameters of the combine harvester according to the feed amount prediction result to achieve real-time control of the feed amount.

[0093] In an embodiment of the present invention, it also includes: a storage unit for storing data of the torque sensor, the speed monitoring module and the feed amount prediction result for subsequent analysis and optimization.

[0094] According to the sensor fusion-based real-time feed amount prediction system for a combine harvester according to an embodiment of the present invention, the torque sensor and speed monitoring module can detect the cross-bridge torque and driving speed of the combine harvester in real time, and establish the state equation, the first observation equation and the second observation equation according to the relationship between the cross-bridge torque, the driving speed and the feed amount, and use the Kalman filter algorithm to perform data fusion to obtain the real-time value of the harvester feed amount. The data after Kalman filtering has a small jump and can accurately follow the changes in the feed amount.

[0095] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A method for real-time feed amount prediction of a combine harvester based on sensor fusion, characterized in that: The steps include: S1. Through experiments, the first observation equation of the cross-bridge torque z1 and the feed amount x is established. According to the definition of the feed amount, the second observation equation of the combine harvester driving speed z2 and the feed amount x is established, and the state equation of the combine harvester torque is established: x t =x t-1 +Δx, where x t is the feed amount of the combine harvester at time t, Δx is the change in the feed amount of the combine harvester from time t-1 to time t; S2, obtaining real-time detection values ​​of the bridge torque and the driving speed; S3, using Kalman filter algorithm to perform data fusion on the detection values ​​of driving speed and bridge torque, so as to realize real-time prediction of the feeding amount of the combine harvester; The step S3 in which the Kalman filter algorithm is used to predict the feed amount of the combine harvester comprises the following steps: S31. Establish a priori estimate of the feed amount of the combine harvester at time K: in, is the prior estimate of the combine harvester feed quantity at time K; is the posterior estimate of the combine harvester feed quantity at time K-1; S32. Establish a priori estimate of the covariance at time K: in, is the prior estimate of the state estimation covariance matrix at time K; P k-1 is the a posteriori estimate of the state estimation covariance matrix at time K-1; Q is the variance of the process noise; S33, update Kalman gain: Among them, K k is the Kalman gain at time K; H is the observation matrix, [H1, H2]; R is the covariance matrix of the observation noise S34, update the posterior estimation of the real-time feeding amount of the combine harvester at time K: in, is the posterior estimation of the real-time feeding amount of the combine harvester at time K; k is the measured value of the feed amount at time K S35. Update the posterior estimate P of the covariance at time K k :

2. The method for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 1, characterized in that: In step S1, the first observation equation is established by using an experimental calibration method, and the establishment method is as follows: Step S11, obtaining the average planting density ρ of the field to be harvested, then the real-time feeding amount x of the combine harvester is: x=ρLV, where V is the travel speed of the combine harvester, and L is the width of the cutting table of the combine harvester; Step S12, make the combine harvester travel a preset distance stably and evenly at a specified feed rate, and record the bridge torque values ​​at different feed rates, calculate the average value μ of the corresponding torque values, and perform linear regression on the data to obtain the first observation equation of the bridge torque z1 and the feed rate x: z1=H1(x)+B Formula 2; Among them, z1 is the measured value of the bridge torque, H1 is the slope of the linear regression, and B is the intercept of the linear regression; Transforming formula 2, we get z'1=z1-B=H1(x) Formula 3; When the influence of the observation noise V1 is added to the measurement, the value of the bridge torque z″1 after adding the observation noise V1 is obtained: z″1=H1(x)+V1 Formula 4; Among them, P(V1)=N(0,R1).

3. The method for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 2, characterized in that: The method for obtaining the average planting density ρ of the field to be harvested in step S11 is: In the field to be harvested, five 1m 3 The straw in the field is taken as the weight of grain, and the grain weight is m, then the average planting density of the field to be harvested is ρ = m / 5.

4. The method for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 1, characterized in that: The method for establishing the second observation equation in step S1 is: According to the real-time feed amount x=ρLV of the combine harvester and the influence of noise V2 in the measurement, the second observation equation of the travel speed z2 and the feed amount x is obtained: z2=H2(x)+V2 Formula 5; Among them, P(V2)=N(0,R2).

5. The method for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 1, characterized in that: The state equation of the torque of the combine harvester established in step S1 is: x t = x t-1 +ρLΔV+W Formula 6; Wherein, P(W)=N(0,Q), W is the process noise, and ΔV is the speed change from time t-1 to time t.

6. A system for real-time feed amount prediction of a combine harvester based on sensor fusion, characterized in that: The system is implemented by the prediction method described in any one of claims 1 to 5, and includes: A torque sensor, wherein the torque sensor is installed on the cross-bridge driving shaft of the combine harvester and is used to measure the torque of the cross-bridge driving shaft of the combine harvester; A measuring unit, the measuring unit being installed in a cab of the combine harvester and comprising a speed monitoring module for measuring a travel speed of the combine harvester; A data processing unit, used to establish a state equation, a first observation equation and a second observation equation according to data from the torque sensor and the measurement unit, and perform data fusion using a Kalman filter algorithm; The output unit is used to output the real-time feed amount prediction result of the combine harvester.

7. The system for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 6, characterized in that: The data processing unit also includes: Mean smoothing filter module, used for smoothing the output data of the torque sensor and speed monitoring module; The Kalman filter module is used to integrate the driving speed and bridge torque data and output accurate feed amount prediction results.

8. The system for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 7, characterized in that: Also includes: The control unit is used to adjust the working parameters of the combine harvester according to the feed amount prediction result to achieve real-time control of the feed amount.

9. The system for real-time feed amount prediction of a combine harvester based on sensor fusion according to claim 7, characterized in that: Also includes: The storage unit is used to store the data of the torque sensor, speed monitoring module and the feed amount prediction results for subsequent analysis and optimization.

Images