Vehicle speed-slip rate combined control method for unmanned rice transplanter

Abstract

The invention relates to a speed-slip rate combined control method for an unmanned rice transplanter, which comprises the following steps: S1, acquiring the speed of the transplanter and the rotating speed of a driving wheel, and calculating the slip rate, speed error and slip rate error of the driving wheel; based on the longitudinal dynamical model of the rice transplanter, adopting Kalman filtering to identify the driving force borne by the driving wheel and the rolling resistance borne by the driving wheel; s2, calculating a rolling resistance change rate according to the speed of the rice transplanter and the rolling resistance borne by the driving wheels obtained in the step S1; s3, based on fuzzy control, a slip rate error proportionality coefficient, a slip rate error integral coefficient, a vehicle speed error proportionality coefficient and a vehicle speed error integral coefficient in a vehicle speed-slip rate combined PI control algorithm are adjusted according to the rolling resistance change rate, the slip rate error and the vehicle speed error; s4, performing combined control on the speed of the rice transplanter and the slip rate of the driving wheel by adopting a speed-slip rate combined PI control algorithm; and S5, repeating the steps S1 to S4.

Description

Technical Field

[0001] The present invention belongs to the field of agricultural automation, and specifically relates to a combined control method for the vehicle speed - slip ratio of an unmanned rice transplanter. Background Technique

[0002] The precise control of intelligent agricultural machinery can effectively improve agricultural production efficiency, enhance operation quality, reduce labor intensity, and solve the problem of labor shortage, which is the future development direction of agricultural machinery. Unmanned rice transplanters are typical representatives of intelligent agricultural machinery, and improving their operation quality is of great significance for increasing China's rice yield. The operation quality of rice transplanters includes row spacing, plant spacing, planting depth, and seedling standing rate, etc. Among them, indicators such as planting depth and seedling standing rate will be affected when the vehicle speed of the transplanter is too high. In terms of plant spacing, due to the fixed transmission ratio between the planting arm and the driving wheel of the transplanter, the plant spacing is affected by the slip rate of the driving wheel. At present, high-speed rice transplanters usually set multiple plant spacing gears to meet the needs of different rice varieties, but the problem of driving wheel slip is not considered in the design process. When the transplanter operates at a constant speed in paddy fields, due to the influence of changes in parameters such as soil parameters, mud depth, and water layer depth, as well as the undulation of the hard bottom layer in paddy fields, the rolling resistance of the driving wheel changes greatly, and the paddy field road surface cannot always provide enough driving force required by the driving wheel, resulting in a large slip rate and its jitter of the driving wheel, which will make it impossible to accurately guarantee the plant spacing of rice transplanting. Therefore, in order to improve the operation quality of rice transplanters, unmanned transplanters need to reduce the slip rate of the driving wheel and improve its stability while ensuring a stable vehicle speed during operation. However, at present, there is only research on controlling the vehicle speed alone in the field of automatic control of transplanters. At the same time, in the fields of road vehicles and agricultural machinery, although there are many related studies on slip rate control, they are not suitable for application to transplanters. Among them, the control objective of the driving wheel slip rate in the field of road vehicles is mainly to improve the vehicle acceleration performance, and the control objective of the driving wheel slip rate in the field of agricultural machinery is mainly to improve the traction utilization efficiency of ploughing agricultural machinery. In the above existing methods, the desired slip rate is usually fixed near the limit slip rate. For rice transplanters, fixing the desired slip rate near the limit slip rate will result in a large gap between the actual plant spacing and the set plant spacing of rice transplanting, which does not meet the requirements of high-quality operation. In addition, the transplanter does not belong to the ploughing agricultural machinery, and the resistance received by its working mechanism is small. Therefore, it is impossible to adjust the driving wheel torque and tillage depth at the same time to control the slip rate and vehicle speed stability like ploughing agricultural machinery. At this time, if the driving wheel slip rate is stabilized near a small fixed value by adjusting the driving wheel torque, due to the strong uncertainty of the paddy field environment, the driving force and resistance received by the whole transplanter cannot be guaranteed to be equal, making it impossible to guarantee the vehicle speed stability. Therefore, in order to ensure the vehicle speed stability, the desired slip rate of the driving wheel of the transplanter will surely jitter within a certain range. For this problem, there are also some vehicle driving anti-skid control methods that only use the driving wheel torque as a single control quantity, judge whether the slip rate exceeds the limit slip rate while performing speed stability control, and if it exceeds the limit slip rate, reduce the driving torque, otherwise, maintain the speed stability control. These methods can keep the driving wheel slip rate within the limit slip rate while ensuring the vehicle speed stability, but the ultimate goal is to prevent the driving wheel from slipping excessively. At this time, the driving wheel slip rate and its jitter are still large.In terms of control algorithms, among all the methods mentioned above, control algorithms that rely on models such as sliding mode control are usually adopted to improve the robustness of the control system. These algorithms require the range of true values of vehicle dynamics model parameters to be known. However, compared with ordinary roads and dry fields, the paddy field environment has greater uncertainties, making it impossible to accurately obtain the range of true values of the dynamics model parameters of the transplanter. Therefore, the existing robust control methods for slip ratio of vehicles and agricultural machinery will face great challenges when applied to rice transplanters. To sum up, at present, there is no reported method with strong robustness that can reduce the slip ratio of the driving wheels of the transplanter while stabilizing its speed and improving the stability of the driving wheel slip ratio. Summary of the Invention

[0003] Aiming at the problem that the operation level of unmanned rice transplanters needs to be broken through and there is a blank in the research on the control of the driving wheel slip ratio, combined with the need to reduce the driving wheel slip ratio and improve the stability of the driving wheel slip ratio during the stable control of the vehicle speed, the purpose of the present invention is to provide a combined control method for the vehicle speed - driving wheel slip ratio of the transplanter. By adjusting the driving torque received by the driving wheels of the transplanter, the vehicle speed and the driving wheel slip ratio of the transplanter are controlled to be stable, and at the same time, the driving wheel slip ratio is reduced to improve the operation quality of the rice transplanter.

[0004] The method of the present invention is mainly divided into two parts. The first part is the identification of the change rate of the rolling resistance in the paddy field. The change rate of the rolling resistance of the driving wheels of the transplanter is identified through Kalman filtering to indirectly reflect the magnitude of the jitter of the paddy field working condition parameters (soil parameters, mud depth, water layer depth, etc.). The second part is the combined control of the vehicle speed - slip ratio based on fuzzy PI control. Fuzzy rules are designed to adjust the control law gain coefficient in real time according to the change rate of the rolling resistance of the driving wheel, the vehicle speed error, and the slip ratio error, so as to achieve strong robust control of the vehicle speed and slip ratio of the transplanter.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A combined control method for the vehicle speed - slip ratio of a transplanter, comprising the following steps:

[0007] S1. Obtain the vehicle speed and the rotational speed of the driving wheels of the transplanter, calculate the driving wheel slip ratio, the vehicle speed error, and the slip ratio error; based on the longitudinal dynamics model of the transplanter, use Kalman filtering to identify the driving force received by the driving wheels and the rolling resistance received by the driving wheels;

[0008] S2. Calculate the change rate of the rolling resistance according to the vehicle speed of the transplanter and the rolling resistance received by the driving wheels obtained in step S1;

[0009] S3. Based on fuzzy control, adjust the proportional coefficient of slip ratio error, integral coefficient of slip ratio error, proportional coefficient of vehicle speed error, and integral coefficient of vehicle speed error in the vehicle speed-slip ratio combined PI control algorithm according to the change rate of rolling resistance, slip ratio error, and vehicle speed error;

[0010] Specifically, it includes the following steps:

[0011] S3.1. Combine the basic domains of each variable and divide the quantization levels of each variable;

[0012] Divide the input variables of vehicle speed error e v , slip ratio error e s , and change rate of rolling resistance τ into three levels: VL(0), M(0.5), and VH(1) respectively. Divide the output variables of proportional coefficient of vehicle speed error K pv , integral coefficient of vehicle speed error K iv , proportional coefficient of slip ratio error K ps , and integral coefficient of slip ratio error K is into five levels: VL(0), ML(0.25), M(0.5), MH(0.75), and VH(1) respectively. Among them, each level of each variable corresponds to a standard Gaussian membership function. The mean value of the membership function of each variable at each level is the value corresponding to the quantization level × (upper bound of the basic domain - lower bound of the basic domain) + lower bound of the basic domain. The standard deviation of the membership function of the input variable at each level is 0.17 × (upper bound of the basic domain - lower bound of the basic domain), and the standard deviation of the membership function of the output variable at each level is 0.106 × (upper bound of the basic domain - lower bound of the basic domain);

[0013] S3.2. Define fuzzy rules;

[0014] Rule 1: If the vehicle speed error e v is at the VL(0) level and the change rate of rolling resistance τ is at the VL(0) level, then the proportional coefficient of vehicle speed error K pv is at the VL(0) level, and the integral coefficient of vehicle speed error K iv is at the M(0.5) level;

[0015] Rule 2: If the vehicle speed error e v is at the VL(0) level and the change rate of rolling resistance τ is at the M(0.5) level, then the proportional coefficient of vehicle speed error K pv is at the ML(0.25) level, and the integral coefficient of vehicle speed error K iv is at the ML(0.25) level;

[0016] Rule 3: If the vehicle speed error e v is at the VL(0) level and the change rate of rolling resistance τ is at the VH(1) level, then the proportional coefficient of vehicle speed error Kpv is of M(0.5) level, and the integral coefficient K of vehicle speed error iv is of VL(0) level;

[0017] Rule 4: If the vehicle speed error e v is of M(0.5) level and the rolling resistance change rate τ is of VL(0) level, then the proportional coefficient K of vehicle speed error pv is of ML(0.25) level, and the integral coefficient K of vehicle speed error iv is of MH(0.75) level;

[0018] Rule 5: If the vehicle speed error e v is of M(0.5) level and the rolling resistance change rate τ is of M(0.5) level, then the proportional coefficient K of vehicle speed error pv is of M(0.5) level, and the integral coefficient K of vehicle speed error iv is of M(0.5) level;

[0019] Rule 6: If the vehicle speed error e v is of M(0.5) level and the rolling resistance change rate τ is of VH(1) level, then the proportional coefficient K of vehicle speed error pv is of MH(0.75) level, and the integral coefficient K of vehicle speed error iv is of ML(0.25) level;

[0020] Rule 7: If the vehicle speed error e v is of VH(1) level and the rolling resistance change rate τ is of VL(0) level, then the proportional coefficient K of vehicle speed error pv is of M(0.5) level, and the integral coefficient K of vehicle speed error iv is of VH(1) level;

[0021] Rule 8: If the vehicle speed error e v is of VH(1) level and the rolling resistance change rate τ is of M(0.5) level, then the proportional coefficient K of vehicle speed error pv is of MH(0.75) level, and the integral coefficient K of vehicle speed error iv is of MH(0.75) level;

[0022] Rule 9: If the vehicle speed error e v is of VH(1) level and the rolling resistance change rate τ is of VH(1) level, then the proportional coefficient K of vehicle speed error pv is of VH(1) level, and the integral coefficient K of vehicle speed error iv is of M(0.5) level;

[0023] Rule 10: If the slip ratio error e sFor the VL(0) level and the rolling resistance change rate τ being at the VL(0) level, the slip rate error proportionality coefficient K ps is at the VL(0) level, and the slip rate error integral coefficient K is is at the M(0.5) level;

[0024] Rule 11: If the slip rate error e s is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip rate error proportionality coefficient K ps is at the ML(0.25) level, and the slip rate error integral coefficient K is is at the ML(0.25) level;

[0025] Rule 12: If the slip rate error e s is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the slip rate error proportionality coefficient K ps is at the M(0.5) level, and the slip rate error integral coefficient K is is at the VL(0) level;

[0026] Rule 13: If the slip rate error e s is at the M(0.5) level and the rolling resistance change rate τ is at the VL(0) level, then the slip rate error proportionality coefficient K ps is at the ML(0.25) level, and the slip rate error integral coefficient K is is at the MH(0.75) level;

[0027] Rule 14: If the slip rate error e s is at the M(0.5) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip rate error proportionality coefficient K ps is at the M(0.5) level, and the slip rate error integral coefficient K is is at the M(0.5) level;

[0028] Rule 15: If the slip rate error e s is at the M(0.5) level and the rolling resistance change rate τ is at the VH(1) level, then the slip rate error proportionality coefficient K ps is at the MH(0.75) level, and the slip rate error integral coefficient K is is at the ML(0.25) level;

[0029] Rule 16: If the slip rate error e s is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the slip rate error proportionality coefficient K ps is at the M(0.5) level, and the slip rate error integral coefficient K is is at the VH(1) level;

[0030] Rule 17. If the slip ratio error e s is at the VH(1) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip ratio error proportionality coefficient K ps is at the MH(0.75) level, and the slip ratio error integral coefficient K is is at the MH(0.75) level;

[0031] Rule 18. If the slip ratio error e s is at the VH(1) level and the rolling resistance change rate τ is at the VH(1) level, then the slip ratio error proportionality coefficient K ps is at the VH(1) level, and the slip ratio error integral coefficient K is is at the M(0.5) level;

[0032] S3.3. Determine the values of each output variable according to the Gaussian membership functions of each input variable at each level and the defined fuzzy rules;

[0033] S3.3.1. Calculate the membership degrees of each input variable at each level according to the Gaussian membership functions of each input variable at each level;

[0034] S3.3.2. Calculate the activation degree of each fuzzy rule according to the membership degrees of each input variable at each level. The activation degree is equal to the minimum value of the membership degrees of all input variables in the corresponding level of this fuzzy rule;

[0035] S3.3.3. Take the activation degree of each fuzzy rule as the membership degree of the output variable in this fuzzy rule; Aggregate the membership degrees of the output variables of all fuzzy rules, that is, obtain the membership degrees of each output variable at each level. If the output variables of multiple fuzzy rules are at the same level, take the maximum membership degree corresponding to this level as the membership degree of this output variable at this level;

[0036] S3.3.4. Multiply the mean value of the membership function of each level by the membership degree of this level, then find the sum of these weighted values, and finally divide by the sum of the membership degrees of all levels;

[0037] S4. Adopt the vehicle speed - slip ratio joint PI control algorithm to jointly control the vehicle speed of the transplanter and the slip ratio of the driving wheel;

[0038] During the control process, judge whether the current slip ratio s of the driving wheel of the transplanter is within the threshold range , represents the upper bound of the expected slip ratio, and s represents the lower bound of the expected slip ratio; If the slip ratio s of the driving wheel is within the threshold range , then the feedforward drive torque T f will accumulate the vehicle speed error e v and the vehicle speed error integral coefficient Kiv The product of, and the vehicle speed error e v and the vehicle speed error proportionality coefficient K pv The product is assigned to the feedback drive torque T b , at this time it belongs to vehicle speed control; conversely, the feedforward drive torque T f Adds the cumulative slip ratio error e s and the slip ratio error integral coefficient K is The product of, and the slip ratio error e s and the slip ratio error proportionality coefficient K ps The product is assigned to the feedback drive torque T b , at this time it belongs to slip ratio control; the sum of the feedforward drive torque T f and the feedback drive torque T b is used as the drive torque T of the drive wheel and output;

[0039] S5. Repeat steps S1 to S4.

[0040] In the said step S1, the calculation formulas of the drive wheel slip ratio, vehicle speed error and slip ratio error are as follows:

[0041] s = 1 - v / (ωr) Formula 1

[0042] e v = v d - v Formula 2

[0043] e s = s d - s Formula 3

[0044] In Formulas 1 to 3, s represents the drive wheel slip ratio; v represents the rice transplanter vehicle speed, with the unit of m·s -1 ; ω represents the drive wheel rotational speed, with the unit of rad·s -1 ; r represents the drive wheel rolling radius, with the unit of m; e v represents the vehicle speed error, with the unit of m·s -1 ; v d represents the desired vehicle speed, with the unit of m·s -1 ; e s represents the slip ratio error; s d represents the desired slip ratio.

[0045] In the said step S1, based on the drive characteristics of the rice transplanter, ignoring the undulations of the hard bottom layer of the paddy field, air resistance, slope resistance, and the relatively small resistance of the working mechanism, a 1 / 4 ideal longitudinal dynamics model of the rice transplanter corresponding to the left rear drive wheel is constructed, and the specific expression is:

[0046]

[0047] In Formula 4, represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the drive wheel, with the unit of rad·s -2 ; F x represents the driving force on the drive wheel, with the unit of N; F r represents the rolling resistance on the drive wheel, with the unit of N; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the drive wheel of the transplanter, with the unit of Kg; T represents the driving torque of the drive wheel, with the unit of N·m; r represents the rolling radius of the drive wheel, with the unit of m; J represents the moment of inertia of the drive wheel, with the unit of Kg·m 2 .

[0048] In the step S1, based on the longitudinal dynamics model of the transplanter, the driving force and rolling resistance on the drive wheel are identified by using the Kalman filter, which specifically includes:

[0049] S1.1. Define the state vector X and the observation vector Z. The state vector The observation vector where v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the drive wheel, with the unit of rad·s -1 ; F x represents the driving force on the drive wheel, with the unit of N; F r represents the rolling resistance on the drive wheel, with the unit of N;

[0050] S1.2. In a single sampling period, it is considered that the driving force F x on the drive wheel and the rolling resistance F r on the drive wheel are basically unchanged. Then the state equation and the observation equation are respectively expressed as:

[0051]

[0052] In Formulas 5 and 6, represents the change rate of the state vector X; represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the drive wheel, with the unit of rad·s -2 ; represents the change rate of the driving force on the drive wheel, with the unit of N·s -1 ; represents the change rate of the rolling resistance on the drive wheel, with the unit of N·s -1 ; F x represents the driving force on the drive wheel, with the unit of N; F r represents the rolling resistance on the drive wheel, with the unit of N; m1 / 4 m represents the mass of the 1 / 4 vehicle model corresponding to the driving wheel of the transplanter, with the unit of Kg; T represents the driving torque of the driving wheel, with the unit of N·m; r represents the rolling radius of the driving wheel, with the unit of m; J represents the moment of inertia of the driving wheel, with the unit of Kg·m 2 ; Z represents the observation vector; X represents the state vector; v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the driving wheel, with the unit of rad·s -1 ;

[0053] S1.3. Let the discrete step size of the system be Δt, and discretize the system to obtain the discrete state equation and the discrete observation equation:

[0054] X k+1 = AX k + BU k + W k Formula 7

[0055] Z k+1 = HX k + V k Formula 8

[0056] In Formula 7 and Formula 8, X k+1 represents the state vector at the (k + 1)-th moment; X k represents the state vector at the k-th moment; U k = T k U k represents the control quantity at the k-th moment, T k represents the driving torque received by the driving wheel at the k-th moment; W k represents the process noise at the k-th moment; A represents the state matrix, B represents the control matrix, where, Δt represents the discrete step size of the system; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the driving wheel of the transplanter, with the unit of Kg; r represents the rolling radius of the driving wheel, with the unit of m; J represents the moment of inertia of the driving wheel, with the unit of Kg·m 2 ; Z k+1 represents the observation vector at the (k + 1)-th moment; Z k represents the observation vector at the k-th moment; V k represents the observation noise at the k-th moment; H represents the observation matrix,

[0057] S1.4. The prediction process of the Kalman filter is as follows:

[0058]

[0059] In Formula 9 and Formula 10, Represents the prior estimated value of the state vector at the k-th moment; Represents the posterior estimated value of the state vector at the (k - 1)-th moment; U k-1 Represents the control quantity at the (k - 1)-th moment; Represents the prior covariance of the error at the k-th moment; P k-1 Represents the covariance of the error at the (k - 1)-th moment; Q represents the covariance matrix of the process noise; A represents the state matrix; B represents the control matrix; A T Represents the transpose matrix of the state matrix;

[0060] S1.5. The update process of the Kalman filter is as follows:

[0061]

[0062] In Formulas 11 to 13, K k Represents the Kalman gain at the k-th moment; Represents the prior covariance of the error at the k-th moment; H represents the observation matrix; H T Represents the transpose matrix of the observation matrix; R represents the covariance matrix of the measurement noise; Represents the posterior estimated value of the state vector at the k-th moment; Represents the prior estimated value of the state vector at the k-th moment; Z k Represents the observation vector at the k-th moment; P k Represents the error at the k-th moment; I represents the identity matrix;

[0063] In the step S1, the speed v of the rice transplanter is measured by the integrated navigation, the rotational speed ω of the drive wheel is measured by the wheel speed sensor, and the driving torque T on the drive wheel is measured by the drive wheel torque sensor.

[0064] In the step S2, the calculation formula for the rolling resistance change rate at the k-th moment is as follows:

[0065]

[0066] In Formula 14, τ k Represents the paddy field rolling resistance change rate at the k-th moment, with the unit of N·mm -1 ; F rk Represents the rolling resistance at the k-th moment, with the unit of N; F r(k-1) Represents the rolling resistance at the (k - 1)-th moment, with the unit of N; v k Represents the speed of the rice transplanter at the k-th moment, with the unit of m·s -1 ; Δt represents the system discrete step length, with the unit of s.

[0067] In the step S3, the form of the standard Gaussian membership function is Among them, μ(x) represents the membership degree, x represents each variable, c represents the mean of the membership function, and σ represents the standard deviation of the membership function.

[0068] The expected slip ratio represents the upper bound of the expected slip ratio, s represents the lower bound of the expected slip ratio, and s represents the driving wheel slip ratio.

[0069] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0070] 1. Improve the operation quality of the unmanned rice transplanter: By stabilizing the vehicle speed of the transplanter while reducing the driving wheel slip ratio and improving the stability of the driving wheel slip ratio, it can effectively ensure the accuracy of the rice planting depth, the seedling standing rate, and the plant spacing accuracy, thereby improving the overall quality of the rice transplanting operation and laying a foundation for high and stable rice yields.

[0071] 2. Adapt to complex working conditions: Identify the change rate of the driving wheel rolling resistance of the transplanter through Kalman filtering to indirectly reflect the jitter magnitude of paddy field working condition parameters (soil parameters, mud depth, water layer depth, etc.). Then design fuzzy rules and adjust the control law gain coefficient in real time according to the driving wheel rolling resistance change rate, vehicle speed error, and slip ratio error to achieve strong robust control of the transplanter vehicle speed and slip ratio.

[0072] 3. Fill the technical gap: At present, there is a lack of research on driving wheel slip ratio control in the field of transplanter automatic control. The present invention fills this key technical gap, provides new ideas and methods for the development of intelligent transplanter control technology, promotes the transplanter automatic control technology to move towards more refined and intelligent directions, and improves the technical level of our country in the field of agricultural intelligent equipment.

[0073] 4. Compatibility and expandability: This control method is essentially a calculation method for the expected driving torque of the driving wheel, with certain compatibility, and can be combined with the hardware system of the existing unmanned rice transplanter, facilitating technical upgrade and transformation. Brief Description of the Drawings

[0074] Figure 1 It is a schematic diagram of the working process of the vehicle speed-slip ratio joint control method of the transplanter of the present invention in each control cycle;

[0075] Figure 2a is the membership function curve diagram of the vehicle speed error e v ;

[0076] Figure 2b is the membership function curve diagram of the slip ratio error e s ;

[0077] Figure 2cIt is the membership function curve graph of the rolling resistance change rate τ;

[0078] Figure 2d is the membership function curve graph of the vehicle speed error proportionality coefficient K pv ;

[0079] Figure 2e is the membership function curve graph of the vehicle speed error integral coefficient K iv ;

[0080] Figure 2f is the membership function curve graph of the slip ratio error proportionality coefficient K ps ;

[0081] Figure 2g is the membership function curve graph of the slip ratio error integral coefficient K is ;

[0082] Figure 3 It is the comparison graph of the vehicle speed and slip ratio of the transplanter under the control of the three methods respectively: the embodiment of the method of the present invention, the pure vehicle speed PI control algorithm (Comparative Example 1), and the vehicle speed - slip ratio joint PI control algorithm in the method of the present invention (Comparative Example 2). At this time, the desired vehicle speed is 0.5 m·s -1 ;

[0083] Figure 4 It is the comparison graph of the vehicle speed and slip ratio of the transplanter under the control of the three methods respectively: the embodiment of the method of the present invention, the pure vehicle speed PI control algorithm (Comparative Example 1), and the vehicle speed - slip ratio joint PI control algorithm in the method of the present invention (Comparative Example 2). At this time, the desired vehicle speed is 0.7 m·s -1 ;

[0084] Figure 5 It is the comparison graph of the vehicle speed and slip ratio of the transplanter under the control of the three methods respectively: the embodiment of the method of the present invention, the pure vehicle speed PI control algorithm (Comparative Example 1), and the vehicle speed - slip ratio joint PI control algorithm in the method of the present invention (Comparative Example 2). At this time, the desired vehicle speed is 1.0 m·s -1 . Specific Embodiments

[0085] The present invention will be further described below with reference to the drawings and embodiments.

[0086] Figure 1 It is the schematic diagram of the working process of the vehicle speed - slip ratio joint control method of the transplanter of the present invention in each control cycle. A vehicle speed - slip ratio joint control method for a transplanter includes the following steps:

[0087] S1. Obtain the vehicle speed and driving wheel rotation speed of the transplanter, calculate the driving wheel slip ratio, vehicle speed error, and slip ratio error; based on the longitudinal dynamics model of the transplanter, use Kalman filtering to identify the driving force and rolling resistance of the driving wheel. This step aims to obtain key information on the current driving state of the transplanter.

[0088] In the above step S1, the calculation formulas for the driving wheel slip ratio, vehicle speed error, and slip ratio error are as follows:

[0089] s = 1 - v / (ωr) Formula 1

[0090] e v = v d - v Formula 2

[0091] e s = s d - s Formula 3

[0092] In Formulas 1 to 3, s represents the driving wheel slip ratio; v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the driving wheel rotation speed, with the unit of rad·s -1 ; r represents the rolling radius of the driving wheel, with the unit of m; e v represents the vehicle speed error, with the unit of m·s -1 ; v d represents the desired vehicle speed, with the unit of m·s -1 ; e s represents the slip ratio error; s d represents the desired slip ratio;

[0093] In the above step S1, based on the driving characteristics of the rice transplanter, ignoring the undulations of the hard bottom layer of the paddy field, air resistance, slope resistance, and the relatively small resistance of the working mechanism, construct a 1 / 4 ideal longitudinal dynamics model of the transplanter corresponding to the left rear driving wheel. The specific expression is:

[0094]

[0095] In Formula 4, represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the driving wheel, with the unit of rad·s -2 ; F x represents the driving force of the driving wheel, with the unit of N; F r represents the rolling resistance of the driving wheel, with the unit of N; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the driving wheel of the transplanter, with the unit of Kg; T represents the driving torque of the driving wheel, with the unit of N·m; r represents the rolling radius of the driving wheel, with the unit of m; J represents the moment of inertia of the driving wheel, with the unit of Kg·m2 ;

[0096] In step S1, based on the longitudinal dynamics model of the transplanter, Kalman filtering is used to identify the driving force on the driving wheel and the rolling resistance on the driving wheel, which specifically includes:

[0097] S1.1. Define the state vector X and the observation vector Z. The state vector The observation vector where v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the driving wheel, with the unit of rad·s -1 ; F x represents the driving force on the driving wheel, with the unit of N; F r represents the rolling resistance on the driving wheel, with the unit of N;

[0098] S1.2. It is considered that the driving force F x on the driving wheel and the rolling resistance F r on the driving wheel are basically unchanged within a single sampling period. Then the state equation and the observation equation are respectively expressed as:

[0099]

[0100] In Formula 5 and Formula 6, represents the change rate of the state vector X; represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the driving wheel, with the unit of rad·s -2 ; represents the change rate of the driving force on the driving wheel, with the unit of N·s -1 ; represents the change rate of the rolling resistance on the driving wheel, with the unit of N·s -1 ; F x represents the driving force on the driving wheel, with the unit of N; F r represents the rolling resistance on the driving wheel, with the unit of N; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the driving wheel of the transplanter, with the unit of Kg; T represents the driving torque of the driving wheel, with the unit of N·m; r represents the rolling radius of the driving wheel, with the unit of m; J represents the moment of inertia of the driving wheel, with the unit of Kg·m 2 ; Z represents the observation vector; X represents the state vector; v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the driving wheel, with the unit of rad·s -1 ;

[0101] S1.3. Assume that the system discrete step size is Δt, and discretize the system to obtain the discrete state equation and the discrete observation equation:

[0102] X k+1 = AX k + BU k + W k Equation 7

[0103] Z k+1 = HX k + V k Equation 8

[0104] In Equations 7 and 8, X k+1 represents the state vector at the (k + 1)-th moment; X k represents the state vector at the k-th moment; U k = T k U k represents the control quantity at the k-th moment, T k represents the driving torque received by the driving wheels at the k-th moment; W k represents the process noise at the k-th moment; A represents the state matrix, B represents the control matrix, where Δt represents the discrete time step of the system; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the driving wheels of the transplanter, with the unit of Kg; r represents the rolling radius of the driving wheels, with the unit of m; J represents the moment of inertia of the driving wheels, with the unit of Kg·m 2 ; Z k+1 represents the observation vector at the (k + 1)-th moment; Z k represents the observation vector at the k-th moment; V k represents the observation noise at the k-th moment; H represents the observation matrix,

[0105] S1.4. The prediction process of the Kalman filter is as follows:

[0106]

[0107] In Equations 9 and 10, represents the prior estimate value of the state vector at the k-th moment; represents the posterior estimate value of the state vector at the (k - 1)-th moment; U k-1 represents the control quantity at the (k - 1)-th moment; represents the prior covariance of the error at the k-th moment; P k-1 represents the covariance of the error at the (k - 1)-th moment; Q represents the covariance matrix of the process noise; A represents the state matrix; B represents the control matrix; A T represents the transpose matrix of the state matrix;

[0108] S1.5. The update process of the Kalman filter is as follows:

[0109]

[0110] In Formulas 11 to 13, K k represents the Kalman gain at the k-th moment; represents the prior covariance of the error at the k-th moment; H represents the observation matrix; H T represents the transpose matrix of the observation matrix; R represents the covariance matrix of the measurement noise; represents the posterior estimate of the state vector at the k-th moment; represents the prior estimate of the state vector at the k-th moment; Z k represents the observation vector at the k-th moment; P k represents the error at the k-th moment; I represents the identity matrix;

[0111] In step S1, the observed value of the speed v of the transplanter can be measured by combined navigation, the observed value of the rotational speed ω of the driving wheel can be measured by a wheel speed sensor, and the driving torque T received by the driving wheel can be measured by a driving wheel torque sensor.

[0112] S2. Calculate the rolling resistance change rate based on the transplanter speed and the rolling resistance received by the driving wheel obtained in step S1, and use this as the jitter degree of the paddy field working condition parameters. This step is used to characterize the uncertainty of the paddy field working condition parameters and provide a basis for adjusting the parameters of the fuzzy controller.

[0113] In step S2, the calculation formula for the rolling resistance change rate at the k-th moment is as follows:

[0114]

[0115] In Formula 14, τ k represents the paddy field rolling resistance change rate at the k-th moment, with the unit of N·mm -1 ; F rk represents the rolling resistance at the k-th moment, with the unit of N; F r(k-1) represents the rolling resistance at the (k - 1)-th moment, with the unit of N; v k represents the transplanter speed at the k-th moment, with the unit of m·s -1 ; Δt represents the system discrete step length, with the unit of s;

[0116] S3. Based on fuzzy control, adjust the proportional coefficient of the slip rate error, the integral coefficient of the slip rate error, the proportional coefficient of the vehicle speed error, and the integral coefficient of the vehicle speed error in the vehicle speed-slip rate joint PI control algorithm according to the rolling resistance change rate (i.e., the jitter degree of the paddy field working condition parameters), the slip rate error, and the vehicle speed error. In this way, the robustness of the control system to the complex paddy field environment is enhanced.

[0117] Specifically, it includes the following steps:

[0118] S3.1. Divide the quantization levels of each variable in combination with the basic domains of the variables;

[0119] Divide the input variable vehicle speed error e v and slip ratio error e s and rolling resistance change rate τ into three levels of VL(0), M(0.5), and VH(1) respectively. Divide the output variable vehicle speed error proportionality coefficient K pv , vehicle speed error integral coefficient K iv , slip ratio error proportionality coefficient K ps and slip ratio error integral coefficient K is into five levels of VL(0), ML(0.25), M(0.5), MH(0.75), and VH(1) respectively, as shown in Table 1 below. Among them, each level of each variable corresponds to a standard Gaussian membership function, and the form of the standard Gaussian membership function is where μ(x) represents the membership degree, x represents each variable, c represents the mean value of the membership function, and σ represents the standard deviation of the membership function. The mean value of the membership function of each variable at each level is the value corresponding to the quantization level × (upper bound of the basic domain - lower bound of the basic domain) + lower bound of the basic domain. The standard deviation of the membership function of the input variable at each level is 0.17 × (upper bound of the basic domain - lower bound of the basic domain), and the standard deviation of the membership function of the output variable at each level is 0.106 × (upper bound of the basic domain - lower bound of the basic domain). For example, the mean value of the Gaussian membership function corresponding to the VH(1) level of the rolling resistance change rate τ = 1×(5 - 0) + 0 = 5, and the standard deviation = 0.1699×(5 - 0) = 0.8495; the mean value of the Gaussian membership function corresponding to the M(0.5) level of the vehicle speed error proportionality coefficient K pv = 0.5×(1200 - 600) + 600 = 900, and the standard deviation = 0.1062×(1200 - 600) = 63.72. The membership function curves of the vehicle speed error e v , slip ratio error e s , rolling resistance change rate τ, vehicle speed error proportionality coefficient K pv , vehicle speed error integral coefficient K iv , slip ratio error proportionality coefficient K ps and slip ratio error integral coefficient K is are as shown in Figures 2a to 2g .

[0120] Table 1 Basic domains and quantization levels of each variable in fuzzy control

[0121]

[0122]

[0123] S3.2. Define fuzzy rules;

[0124] Rule 1: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the VL(0) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level;

[0125] Rule 2: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K iv is at the ML(0.25) level;

[0126] Rule 3: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VL(0) level;

[0127] Rule 4: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K iv is at the MH(0.75) level;

[0128] Rule 5: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level;

[0129] Rule 6: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the MH(0.75) level, and the vehicle speed error integral coefficient K iv is at the ML(0.25) level;

[0130] Rule 7: If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pvis at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VH(1) level;

[0131] Rule 8: If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the MH(0.75) level, and the vehicle speed error integral coefficient K iv is at the MH(0.75) level;

[0132] Rule 9: If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the VH(1) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level;

[0133] Rule 10: If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the VL(0) level, then the slip ratio error proportionality coefficient K ps is at the VL(0) level, and the slip ratio error integral coefficient K is is at the M(0.5) level;

[0134] Rule 11: If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip ratio error proportionality coefficient K ps is at the ML(0.25) level, and the slip ratio error integral coefficient K is is at the ML(0.25) level;

[0135] Rule 12: If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the slip ratio error proportionality coefficient K ps is at the M(0.5) level, and the slip ratio error integral coefficient K is is at the VL(0) level;

[0136] Rule 13: If the slip ratio error e s is at the M(0.5) level and the rolling resistance change rate τ is at the VL(0) level, then the slip ratio error proportionality coefficient K ps is at the ML(0.25) level, and the slip ratio error integral coefficient K is is at the MH(0.75) level;

[0137] Rule 14: If the slip ratio error e sFor the M(0.5) level and the rolling resistance change rate τ being at the M(0.5) level, the slip rate error proportionality coefficient K ps is at the M(0.5) level, and the slip rate error integral coefficient K is is at the M(0.5) level;

[0138] Rule 15: If the slip rate error e s is at the M(0.5) level and the rolling resistance change rate τ is at the VH(1) level, then the slip rate error proportionality coefficient K ps is at the MH(0.75) level, and the slip rate error integral coefficient K is is at the ML(0.25) level;

[0139] Rule 16: If the slip rate error e s is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the slip rate error proportionality coefficient K ps is at the M(0.5) level, and the slip rate error integral coefficient K is is at the VH(1) level;

[0140] Rule 17: If the slip rate error e s is at the VH(1) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip rate error proportionality coefficient K ps is at the MH(0.75) level, and the slip rate error integral coefficient K is is at the MH(0.75) level;

[0141] Rule 18: If the slip rate error e s is at the VH(1) level and the rolling resistance change rate τ is at the VH(1) level, then the slip rate error proportionality coefficient K ps is at the VH(1) level, and the slip rate error integral coefficient K is is at the M(0.5) level;

[0142] S3.3. Determine the values of each output variable according to the Gaussian membership functions of each input variable at each level and the defined fuzzy rules;

[0143] S3.3.1. Calculate the membership degrees of each input variable at each level according to the Gaussian membership functions of each input variable at each level;

[0144] S3.3.2. Calculate the activation degree of each fuzzy rule according to the membership degrees of each input variable at each level. The activation degree is equal to the minimum value of the membership degrees of all input variables in the corresponding level of this fuzzy rule (i.e., the activation degree of the fuzzy rule is determined by the membership degrees of all input variables in the corresponding level, and the minimum value is selected as the activation degree);

[0145] S3.3.3. Take the activation degree of each fuzzy rule as the membership degree of the output variable in that fuzzy rule; aggregate the membership degrees of the output variables of all fuzzy rules, that is, obtain the membership degree of each output variable at each level. If the output variable of multiple fuzzy rules is at the same level, take the maximum membership degree corresponding to that level as the membership degree of the output variable at that level.

[0146] S3.3.4. Calculate the value of the output variable. The value of the output variable is obtained by weighted calculation of the membership degree corresponding to each level. The specific method is to multiply the mean value of the membership degree function of each level by the membership degree of that level, then find the sum of these weighted values, and finally divide by the sum of the membership degrees of all levels.

[0147] Take the vehicle speed error e v and the rolling resistance change rate τ to calculate the vehicle speed error proportionality coefficient K pv and the vehicle speed error integral coefficient K iv as an example. Assume the current vehicle speed error e v = 0.1 m·s -1 , and the rolling resistance change rate τ = 1 N·mm -1 .

[0148] First, according to the vehicle speed error e v and the value of the rolling resistance change rate τ, as well as the Gaussian membership degree functions of both at each level, calculate the membership degrees of both at each level:

[0149] For the vehicle speed error e v , at the VL(0) level, at the M(0.5) level, at the VH(1) level,

[0150] For the rolling resistance change rate τ, at the VL(0) level, at the M(0.5) level, at the VH(1) level,

[0151] Next, according to the membership degrees of each input variable at each level, calculate the activation degree of each fuzzy rule. The activation degree is equal to the minimum value of the membership degrees of all input variables in the fuzzy rule corresponding to that level. At the same time, calculate the membership degree of the output variable in each rule corresponding to that level in that rule, which is equal to the activation degree of that rule. The detailed calculation process is as follows:

[0152] Rule 1 is defined as: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pvis at the VL(0) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level. The activation degree of this fuzzy rule = min(the membership degree of the vehicle speed error e v at the VL(0) level, the membership degree of the rolling resistance change rate τ at the VL(0) level) = min(0.84, 0.50) = 0.50, and the vehicle speed error ratio coefficient K pv at the VL(0) level is the membership degree of the vehicle speed error integral coefficient K iv at the M(0.5) level is 0.50.

[0153] Rule 2 is defined as: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error ratio coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K iv is at the ML(0.25) level. The activation degree of this fuzzy rule = min(the membership degree of the vehicle speed error e v at the VL(0) level, the membership degree of the rolling resistance change rate τ at the M(0.5) level) = min(0.84, 0.21) = 0.21, and the vehicle speed error ratio coefficient K pv at the ML(0.25) level is the membership degree of the vehicle speed error integral coefficient K iv at the ML(0.25) level is 0.21.

[0154] Rule 3 is defined as: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error ratio coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VL(0) level. The activation degree of this fuzzy rule = min(the membership degree of the vehicle speed error e v at the VL(0) level, the membership degree of the rolling resistance change rate τ at the VH(1) level) = min(0.84, 0.00) = 0.00, and the vehicle speed error ratio coefficient K pv at the M(0.5) level is the membership degree of the vehicle speed error integral coefficient K iv at the VL(0) level is 0.00.

[0155] Rule 4 is defined as: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error ratio coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K ivis at the MH(0.75) level. The activation degree of this fuzzy rule = min(vehicle speed error e v is the membership degree at the M(0.5) level, and the rolling resistance change rate τ is the membership degree at the VL(0) level) = min(0.06, 0.50) = 0.06. The vehicle speed error proportionality coefficient K pv is the membership degree at the ML(0.25) level = the vehicle speed error integral coefficient K iv is the membership degree at the MH(0.75) level = 0.06.

[0156] Rule 5 is defined as: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level. The activation degree of this fuzzy rule = min(vehicle speed error e v is the membership degree at the M(0.5) level, and the rolling resistance change rate τ is the membership degree at the M(0.5) level) = min(0.06, 0.21) = 0.06. The membership degree of K pv at the M(0.5) level = the membership degree of the vehicle speed error integral coefficient K iv at the M(0.5) level = 0.06.

[0157] Rule 6 is defined as: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the MH(0.75) level, and the vehicle speed error integral coefficient K iv is at the ML(0.25) level. The activation degree of this fuzzy rule = min(vehicle speed error e v is the membership degree at the M(0.5) level, and the rolling resistance change rate τ is the membership degree at the VH(1) level) = min(0.06, 0.00) = 0.00. The membership degree of the vehicle speed error proportionality coefficient K pv at the MH(0.75) level = the membership degree of the vehicle speed error integral coefficient K iv at the ML(0.25) level = 0.00.

[0158] Rule 7 is defined as: If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VH(1) level. The activation degree of this fuzzy rule = min(vehicle speed error e vis the membership degree of VH(1) level, and the rolling resistance change rate τ is the membership degree of VL(0) level) = min(0.00, 0.50) = 0.00, the vehicle speed error proportionality coefficient K pv is the membership degree of M(0.5) level = the vehicle speed error integral coefficient K iv is the membership degree of VH(1) level = 0.00.

[0159] Rule 8 is defined as: If the vehicle speed error e v is of VH(1) level and the rolling resistance change rate τ is of M(0.5) level, then the vehicle speed error proportionality coefficient K pv is of MH(0.75) level, and the vehicle speed error integral coefficient K iv is of MH(0.75) level. The activation degree of this fuzzy rule = min(the membership degree of vehicle speed error e v is of VH(1) level, the membership degree of rolling resistance change rate τ is of M(0.5) level) = min(0.00, 0.21) = 0.00, the vehicle speed error proportionality coefficient K pv is the membership degree of MH(0.75) level = the vehicle speed error integral coefficient K iv is the membership degree of MH(0.75) level = 0.00.

[0160] Rule 9 is defined as: If the vehicle speed error e v is of VH(1) level and the rolling resistance change rate τ is of VH(1) level, then the vehicle speed error proportionality coefficient K pv is of VH(1) level, and the vehicle speed error integral coefficient K iv is of M(0.5) level. The activation degree of this fuzzy rule = min(the membership degree of vehicle speed error e v is of VH(1) level, the membership degree of rolling resistance change rate τ is of VH(1) level) = min(0.00, 0.00) = 0.00, the vehicle speed error proportionality coefficient K pv is the membership degree of VH(1) level = the vehicle speed error integral coefficient K iv is the membership degree of M(0.5) level = 0.00.

[0161] Then, aggregate the membership degrees of all output variables, that is, obtain the membership degree of each output variable at each level. If an output variable of multiple fuzzy rules is at the same level, take the maximum membership degree corresponding to this level as the membership degree of this output variable at this level; the specific process is as follows:

[0162] The vehicle speed error proportionality coefficient K pv is the membership degree of VL(0) level = the activation degree of Rule 1 = 0.50; the vehicle speed error proportionality coefficient K pvThe membership degree for the ML(0.25) level = max(activation degree of Rule 2, activation degree of Rule 4) = max(0.21, 0.06) = 0.21; vehicle speed error proportionality coefficient K pv The membership degree for the M(0.5) level = max(activation degree of Rule 3, activation degree of Rule 5, activation degree of Rule 7) = max(0.00, 0.06, 0.00) = 0.06; vehicle speed error proportionality coefficient K pv The membership degree for the MH(0.75) level = max(activation degree of Rule 6, activation degree of Rule 8) = max(0.00, 0.00) = 0.00; vehicle speed error proportionality coefficient K pv The membership degree for the VH(1) level = activation degree of Rule 9 = 0.00.

[0163] Vehicle speed error integral coefficient K iv The membership degree for the VL(0) level = activation degree of Rule 3 = 0.00; vehicle speed error integral coefficient K iv The membership degree for the ML(0.25) level = max(activation degree of Rule 2, activation degree of Rule 6) = max(0.21, 0.00) = 0.21; vehicle speed error integral coefficient K iv The membership degree for the M(0.5) level = max(activation degree of Rule 1, activation degree of Rule 5, activation degree of Rule 9) = max(0.50, 0.06, 0.00) = 0.50; vehicle speed error integral coefficient K iv The membership degree for the MH(0.75) level = max(activation degree of Rule 4, activation degree of Rule 8) = max(0.06, 0.00) = 0.06; vehicle speed error integral coefficient K iv The membership degree for the VH(1) level = activation degree of Rule 7 = 0.00.

[0164] Finally, calculate the value of the output variable. The value of the output variable is obtained by weighted calculation of the membership degree corresponding to each level. The specific method is to multiply the mean value of the membership degree function of each level by the membership degree of that level, then find the sum of these weighted values, and finally divide by the sum of the membership degrees of all levels. The specific process is as follows:

[0165] Similarly, according to the slip ratio error e s and the value of the rolling resistance change rate τ, calculate the slip ratio error proportionality coefficient K ps and the slip ratio error integral coefficient K is .

[0166] S4. Adopt the vehicle speed - slip ratio combined PI control algorithm to jointly control the vehicle speed of the transplanter and the driving wheel slip ratio;

[0167] Before joint control, first specify the threshold range of the driving wheel slip ratio s represents the upper bound of the desired slip ratio, s represents the lower bound of the desired slip ratio; where, s and should both be small to ensure that the slip ratio is at a small value and varies within a small range.

[0168] During the control process, determine whether the current driving wheel slip ratio s of the transplanter is within the threshold range ; if the driving wheel slip ratio s is within the threshold range , then the feedforward driving torque T f will accumulate the product of the vehicle speed error e v and the vehicle speed error integral coefficient K iv , and assign the product of the vehicle speed error e v and the vehicle speed error proportional coefficient K pv to the feedback driving torque T b . At this time, it belongs to vehicle speed control; conversely, the feedforward driving torque T f will accumulate the product of the slip ratio error e s and the slip ratio error integral coefficient K is , and assign the product of the slip ratio error e s and the slip ratio error proportional coefficient K ps to the feedback driving torque T b . At this time, it belongs to slip ratio control; the sum of the feedforward driving torque T f and the feedback driving torque T b is used as the driving torque T of the driving wheel and output. It should be noted that the desired slip ratio in the present invention is to ensure that during the slip ratio control process, even if the slip ratio enters the threshold range, there is still a tendency to change towards the center of the threshold range, so as to improve the stability of the slip ratio.

[0169] S5. Repeat steps S1 to S4. In this way, the joint control of the vehicle speed and slip ratio of the transplanter is achieved.

[0170] Embodiment

[0171] In this embodiment, the rice transplanter used in the experiment is an electric transplanter, and the gasoline engine in its drive system is replaced by an AC induction motor, and its controller can directly control its output torque. Therefore, the output torque of the drive motor can be determined according to the transmission ratio between the drive motor and the drive wheel and the drive torque of the drive wheel calculated by the method of the present invention, and then the drive wheel can obtain the desired drive torque.

[0172] Figure 3 、 Figure 4 、Figure 5 respectively for the expected vehicle speeds of 0.5 m·s -1 , 0.7 m·s -1 , 1 m·s -1 when, the comparison charts of the vehicle speed and slip ratio of the rice transplanter under the embodiment of the method of the present invention and Comparative Example 1 (using a pure vehicle speed PI control algorithm) and Comparative Example 2 (directly using the vehicle speed-slip ratio combined PI control algorithm in the method of the present invention). Table 2 is the statistical results of the vehicle speed and slip ratio control under the above control methods. During the experiment, it is stipulated that s = 0, s d = 0.05 to ensure that the slip ratio is at a small value and varies within a small range. In the pure vehicle speed PI control algorithm, K pv = 900, K iv = 450. In the vehicle speed-slip ratio combined control algorithm based on PI control, K pv = 900, K ps = 9000, K iv = 450, K is = 4500. In the Kalman filtering process of the method of the present invention, Q = diag(0.01, 0.05, 25, 25), R = diag(0.0004, 0.0064). Among them, diag(*) represents a diagonal matrix. The initial value of the state vector X0 = [0, 0, 500, 200] T , and the initial value of the error covariance matrix P0 = diag(0.01, 0.05, 25, 25). The control period is 0.2 s.

[0173] Taking v d = 1 m / s as an example, the control results of the slip ratio and vehicle speed under different algorithms are compared and analyzed. In terms of slip ratio control, the mean value of the slip ratio under the control of the method of the present invention is 0.122, the standard deviation is 0.039, and the proportion greater than 0.1 is 70.1%; compared with no slip ratio control (the pure vehicle speed PI control algorithm of Comparative Example 1), the mean value of the slip ratio is reduced by 39.3%, the standard deviation is reduced by 61.7%, and the proportion greater than 0.1 is reduced by 21.2%; compared with the vehicle speed-slip ratio combined PI control algorithm with fixed control law gain coefficients, the mean value of the slip ratio is reduced by 12.8%, the standard deviation is reduced by 45.1%, and the proportion greater than 0.1 is reduced by 6.7%. In terms of vehicle speed control, the mean value of the absolute error of the vehicle speed under the control of the method of the present invention is 0.072 m·s -1, it is reduced by 53.5% compared with the situation without slip ratio control, and is reduced by 47.1% compared with the proportional control example 2 vehicle speed-slip ratio combined PI control algorithm with fixed control law gain coefficient. It can be shown from this that the method of the present invention can effectively reduce the driving wheel slip ratio and its jitter while ensuring the stability of the rice transplanter speed, and obtains strong robustness by identifying the rolling resistance change rate and fuzzy control, and can improve the precision and quality of rice transplanting.

[0174] Table 2

[0175]

Claims

1. A combined control method for the vehicle speed - slip ratio of a transplanter, characterized in that, The combined control method for the vehicle speed - slip ratio of the transplanter includes the following steps: S1. Obtain the vehicle speed and the rotational speed of the driving wheels of the transplanter, calculate the slip ratio of the driving wheels, the vehicle speed error, and the slip ratio error; based on the longitudinal dynamics model of the transplanter, use Kalman filtering to identify the driving force and the rolling resistance acting on the driving wheels; S2. Calculate the change rate of the rolling resistance according to the vehicle speed of the transplanter and the rolling resistance acting on the driving wheels obtained in step S1; S3. Based on fuzzy control, adjust the proportional coefficient of the slip ratio error, the integral coefficient of the slip ratio error, the proportional coefficient of the vehicle speed error, and the integral coefficient of the vehicle speed error in the combined PI control algorithm for the vehicle speed - slip ratio according to the change rate of the rolling resistance, the slip ratio error, and the vehicle speed error; Specifically, it includes the following steps: S3.

1. Combine the basic domains of each variable and divide the quantization levels of each variable; The input variable vehicle speed error e v , slip ratio error e s and rolling resistance change rate τ are respectively divided into three levels: VL(0), M(0.5), and VH(1). The output variable vehicle speed error proportionality coefficient K pv , vehicle speed error integral coefficient K iv , slip ratio error proportionality coefficient K ps and slip ratio error integral coefficient K is are respectively divided into five levels: VL(0), ML(0.25), M(0.5), MH(0.75), and VH(1). Among them, each level of each variable corresponds to a standard Gaussian membership function. The mean value of the membership function of each variable at each level is the value corresponding to the quantization level × (upper bound of the basic domain - lower bound of the basic domain) + lower bound of the basic domain. The standard deviation of the membership function of the input variable at each level is 0.17 × (upper bound of the basic domain - lower bound of the basic domain), and the standard deviation of the membership function of the output variable at each level is 0.106 × (upper bound of the basic domain - lower bound of the basic domain); S3.

2. Define the fuzzy rules; Rule 1: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the VL(0) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level; Rule 2: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K iv is at the ML(0.25) level; Rule 3: If the vehicle speed error e v is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VL(0) level; Rule 4: If the vehicle speed error e v is at the M(0.5) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the ML(0.25) level, and the vehicle speed error integral coefficient K iv is at the MH(0.75) level; Rule 5. If the vehicle speed error e v is at M(0.5) level and the rolling resistance change rate τ is at M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at M(0.5) level, and the vehicle speed error integral coefficient K iv is at M(0.5) level; Rule 6. If the vehicle speed error e v is at M(0.5) level and the rolling resistance change rate τ is at VH(1) level, then the vehicle speed error proportionality coefficient K pv is at MH(0.75) level, and the vehicle speed error integral coefficient K iv is at ML(0.25) level; Rule 7. If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the vehicle speed error proportionality coefficient K pv is at the M(0.5) level, and the vehicle speed error integral coefficient K iv is at the VH(1) level; Rule 8. If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the M(0.5) level, then the vehicle speed error proportionality coefficient K pv is at the MH(0.75) level, and the vehicle speed error integral coefficient K iv is at the MH(0.75) level; Rule 9. If the vehicle speed error e v is at the VH(1) level and the rolling resistance change rate τ is at the VH(1) level, then the vehicle speed error proportionality coefficient K pv is at the VH(1) level, and the vehicle speed error integral coefficient K iv is at the M(0.5) level; Rule 10. If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the VL(0) level, then the slip ratio error proportionality coefficient K ps is at the VL(0) level, and the slip ratio error integral coefficient K is is at the M(0.5) level; Rule 11. If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip ratio error proportionality coefficient K ps is at the ML(0.25) level, and the slip ratio error integral coefficient K is is at the ML(0.25) level; Rule 12. If the slip ratio error e s is at the VL(0) level and the rolling resistance change rate τ is at the VH(1) level, then the slip ratio error proportionality coefficient K ps is at the M(0.5) level, and the slip ratio error integral coefficient K is is at the VL(0) level; Rule 13. If the slip ratio error e s is at M(0.5) level and the rolling resistance change rate τ is at VL(0) level, then the slip ratio error proportionality coefficient K ps is at ML(0.25) level, and the slip ratio error integral coefficient K is is at MH(0.75) level; Rule 14. If the slip ratio error e s is at M(0.5) level and the rolling resistance change rate τ is at M(0.5) level, then the slip ratio error proportionality coefficient K ps is at M(0.5) level, and the slip ratio error integral coefficient K is is at M(0.5) level; Rule 15. If the slip ratio error e s is at M(0.5) level and the rolling resistance change rate τ is at VH(1) level, then the slip ratio error proportionality coefficient K ps is at MH(0.75) level, and the slip ratio error integral coefficient K is is at ML(0.25) level; Rule 16. If the slip ratio error e s is at the VH(1) level and the rolling resistance change rate τ is at the VL(0) level, then the slip ratio error proportionality coefficient K ps is at the M(0.5) level, and the slip ratio error integral coefficient K is is at the VH(1) level; Rule 17. If the slip ratio error e s is at the VH(1) level and the rolling resistance change rate τ is at the M(0.5) level, then the slip ratio error proportionality coefficient K ps is at the MH(0.75) level, and the slip ratio error integral coefficient K is is at the MH(0.75) level; Rule 18. If the slip ratio error e s is at the VH(1) level and the rolling resistance change rate τ is at the VH(1) level, then the slip ratio error proportionality coefficient K ps is at the VH(1) level, and the slip ratio error integral coefficient K is is at the M(0.5) level; S3.

3. Determine the values of each output variable according to the Gaussian membership functions of each input variable at each level and the defined fuzzy rules; S3.3.

1. Calculate the membership degree of each input variable at each level according to the Gaussian membership functions of each input variable at each level; S3.3.

2. Calculate the activation degree of each fuzzy rule. The activation degree is equal to the minimum value of the membership degrees of all input variables in the corresponding level of this fuzzy rule; S3.3.

3. Take the activation degree of each fuzzy rule as the membership degree of the output variable in this fuzzy rule; aggregate the membership degrees of the output variables of all fuzzy rules, that is, obtain the membership degree of each output variable at each level. If the output variables of multiple fuzzy rules are at the same level, take the maximum membership degree corresponding to this level as the membership degree of the output variable at this level; S3.3.

4. Multiply the mean value of the membership function of each level by the membership degree of this level, then calculate the sum of these weighted values, and finally divide by the sum of the membership degrees of all levels; S4. Adopt the combined PI control algorithm for the vehicle speed - slip ratio to jointly control the vehicle speed and the slip ratio of the driving wheels of the transplanter; During the control process, it is judged whether the current slip ratio s of the driving wheel of the transplanter is within the threshold range Inside, represents the upper bound of the expected slip ratio, and s represents the lower bound of the expected slip ratio; if the slip ratio s of the driving wheel is within the threshold range Inside, then the feedforward driving torque T f Will accumulate the vehicle speed error e v The product with the vehicle speed error integral coefficient K iv And the vehicle speed error e v The product with the vehicle speed error proportionality coefficient K pv Is assigned to the feedback driving torque T b , At this time, it belongs to vehicle speed control; on the contrary, the feedforward driving torque T f Will accumulate the slip ratio error e s The product with the slip ratio error integral coefficient K is And the slip ratio error e s The product with the slip ratio error proportionality coefficient K ps Is assigned to the feedback driving torque T b , At this time, it belongs to slip ratio control; the sum of the feedforward driving torque T f And the feedback driving torque T b Is used as the driving torque T of the driving wheel and output; S5. Repeat steps S1 to S4.

2. The combined control method of the speed-slip ratio of the rice transplanter according to claim 1, characterized in that In step S1, the calculation formulas for the slip ratio of the driving wheels, the vehicle speed error, and the slip ratio error are as follows: s = 1 - v / (ωr) Formula 1e v = v d -v Formula 2e s = s d -s Formula 3 In Formulas 1 to 3, s represents the driving wheel slip ratio; v represents the speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the driving wheel, with the unit of rad·s -1 ; r represents the rolling radius of the driving wheel, with the unit of m; e v represents the speed error, with the unit of m·s -1 ; v d represents the desired speed, with the unit of m·s -1 ; e s represents the slip ratio error; s d represents the desired slip ratio.

3. The combined control method of the speed - slip ratio of the transplanter according to claim 1, wherein In step S1, based on the driving characteristics of the rice transplanter, ignoring the undulations of the hard bottom layer of the paddy field, air resistance, slope resistance, and the relatively small resistance of the working mechanism, construct a 1 / 4 ideal longitudinal dynamics model of the transplanter corresponding to the left - rear driving wheel. The specific expression is: In Formula 4, represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the drive wheel, with the unit of rad·s -2 ; F x represents the driving force on the drive wheel, with the unit of N; F r represents the rolling resistance on the drive wheel, with the unit of N; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the drive wheel of the transplanter, with the unit of Kg; T represents the driving torque of the drive wheel, with the unit of N·m; r represents the rolling radius of the drive wheel, with the unit of m; J represents the moment of inertia of the drive wheel, with the unit of Kg·m 2 .

4. The joint control method of the speed-slip ratio of the transplanter according to claim 1, characterized in that In step S1, based on the longitudinal dynamics model of the transplanter, use Kalman filtering to identify the driving force and the rolling resistance acting on the driving wheels. Specifically, it includes: S1.

1. Define the state vector X and the observation vector Z. The state vector The observation vector where v represents the speed of the transplanter, in m·s -1 ; ω represents the rotational speed of the driving wheel, in rad·s -1 ; F x represents the driving force on the driving wheel, in N; F r represents the rolling resistance on the driving wheel, in N; S1.

2. During a single sampling period, it is considered that the driving force F x acting on the driving wheel and the rolling resistance F r acting on the driving wheel are basically unchanged. Then, the state equation and the observation equation are respectively expressed as: In Formula 5 and Formula 6, represents the change rate of the state vector X; represents the acceleration of the transplanter, with the unit of m·s -2 ; represents the angular acceleration of the drive wheel, with the unit of rad·s -2 ; represents the change rate of the driving force received by the drive wheel, with the unit of N·s -1 ; represents the change rate of the rolling resistance received by the drive wheel, with the unit of N·s -1 ; F x represents the driving force received by the drive wheel, with the unit of N; F r represents the rolling resistance received by the drive wheel, with the unit of N; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the drive wheel of the transplanter, with the unit of Kg; T represents the driving torque of the drive wheel, with the unit of N·m; r represents the rolling radius of the drive wheel, with the unit of m; J represents the moment of inertia of the drive wheel, with the unit of Kg·m 2 ; Z represents the observation vector; X represents the state vector; v represents the vehicle speed of the transplanter, with the unit of m·s -1 ; ω represents the rotational speed of the drive wheel, with the unit of rad·s -1 ; S1.

3. Assume that the system discrete step size is Δt, and discretize the system to obtain the discrete state equation and the discrete observation equation: X k+1 = AX k + BU k + W k Equation 7Z k+1 = HX k + V k In Equation 7 and Equation 8, X k+1 represents the state vector at the (k + 1)-th moment; X k represents the state vector at the k-th moment; U k = T k , U k represents the control quantity at the k-th moment, T k represents the driving torque on the driving wheel at the k-th moment; W k represents the process noise at the k-th moment; A represents the state matrix, B represents the control matrix, where Δt represents the discrete step length of the system; m 1 / 4 represents the mass of the 1 / 4 vehicle model corresponding to the driving wheel of the transplanter, with the unit of Kg; r represents the rolling radius of the driving wheel, with the unit of m; J represents the moment of inertia of the driving wheel, with the unit of Kg·m 2 ; Z k+1 represents the observation vector at the (k + 1)-th moment; Z k represents the observation vector at the k-th moment; V k represents the observation noise at the k-th moment; H represents the observation matrix, S1.

4. The prediction process of Kalman filtering is: In Equation 9 and Equation 10, represents the prior estimate of the state vector at the k-th moment; represents the posterior estimate of the state vector at the (k - 1)-th moment; U k-1 represents the control quantity at the (k - 1)-th moment; represents the prior covariance of the error at the k-th moment; P k-1 represents the covariance of the error at the (k - 1)-th moment; Q represents the covariance matrix of the process noise; A represents the state matrix; B represents the control matrix; A T represents the transpose matrix of the state matrix; S1.

5. The update process of Kalman filtering is: In Formulas 11 to 13, K k represents the Kalman gain at the k-th moment; represents the prior covariance of the error at the k-th moment; H represents the observation matrix; H T represents the transpose matrix of the observation matrix; R represents the covariance matrix of the measurement noise; represents the posterior estimate of the state vector at the k-th moment; represents the prior estimate of the state vector at the k-th moment; Z k represents the observation vector at the k-th moment; P k represents the error at the k-th moment; I represents the identity matrix.

5. The combined control method of the speed - slip ratio of the rice transplanter according to claim 1, characterized in that, In step S1, the vehicle speed v of the transplanter is measured by combined navigation, the rotational speed ω of the driving wheels is measured by a wheel speed sensor, and the driving torque T acting on the driving wheels is measured by a driving wheel torque sensor.

6. The combined control method of the speed-slip rate of the rice transplanter according to claim 1, characterized in that, In the step S2, the calculation formula for the rolling resistance change rate at the k-th moment is as follows: In Formula 14, τ k represents the change rate of the paddy field rolling resistance at the k-th moment, with the unit of N·mm -1 ; F rk represents the rolling resistance at the k-th moment, with the unit of N; F r(k-1) represents the rolling resistance at the (k - 1)-th moment, with the unit of N; v k represents the speed of the transplanter at the k-th moment, with the unit of m·s -1 ; Δt represents the discrete step length of the system, with the unit of s.

7. The combined control method of the speed - slip ratio of the rice transplanter according to claim 1, characterized in that, In the step S3, the form of the standard Gaussian membership function is where μ(x) represents the membership degree, x represents each variable, c represents the mean of the membership function, and σ represents the standard deviation of the membership function.

8. The combined control method of the speed - slip ratio of the rice transplanter according to claim 2, characterized in that, The desired slip ratio represents the upper bound of the desired slip ratio, s represents the lower bound of the desired slip ratio, and s represents the driving wheel slip ratio.

Images