PID-DOB composite driving track control method for hydraulic mechanical arm

Abstract

The invention relates to a PID-DOB composite driving track control method for a hydraulic mechanical arm, and belongs to the field of track gauges of hydraulic mechanical arms. The method comprises the steps that firstly, modeling is conducted on a hydraulic system of the hydraulic mechanical arm; the saturation characteristic and the dead zone characteristic of the hydraulic system are analyzed; secondly, all shafts of the mechanical arm are identified and modeled by analyzing a signal spectrogram of normal work of the hydraulic mechanical arm, second-order linear system models of all the shafts are obtained, and a segmented PID control method is designed on the basis of the identified second-order linear system models; composite PID-DOB control rate design is carried out on the basis of a PID control method; the DOB estimates and compensates external disturbance and modeling errors of the system in real time; according to the segmented PID control method, the control process is divided into a starting segment, a main body segment and a braking segment according to time, speed and position information. The control precision and stability of the hydraulic mechanical arm are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of trajectory regulation of hydraulic manipulators, and relates to a PID-DOB composite driving high-precision trajectory control method for hydraulic manipulators. BACKGROUND

[0002] Hydraulic manipulators play a vital role as core equipment in key fields such as industrial automation, aerospace, and ocean engineering. However, a hydraulic manipulator system is essentially a complex nonlinear dynamic system. Its strong nonlinearity is mainly derived from multiple aspects, including the inherent compressibility of hydraulic oil, the nonlinear flow characteristics of control valve ports, and the friction existing in the system. These factors are coupled with each other, bringing great challenges to the accurate mathematical modeling and high-performance control of the system.

[0003] In addition, hydraulic manipulators also face complex uncertainty problems in actual operation. These uncertainties are specifically manifested in the following aspects: the moment of inertia of the system may change with the load, the friction coefficients of each joint are difficult to accurately measure, the bulk modulus of the hydraulic oil fluctuates with temperature and pressure, and the flow coefficient of the servo valve may also drift. These complex uncertainties with unknown boundaries will directly and severely affect the trajectory following accuracy of the manipulator end effector and the overall operation stability of the system.

[0004] To solve the above problems, researchers have proposed various control strategies. The traditional proportional-integral-derivative (PID) control is widely used due to its simple structure and ease of implementation. However, the PID controller is not capable of handling complex uncertainties such as rapid time-varying and unknown boundaries, and it is difficult to meet the stringent requirements of hydraulic manipulators in high-precision trajectory following applications.

[0005] Model predictive control (MPC) is another advanced control method, but its performance is highly dependent on an accurate system model. Given the inherent complexity of hydraulic manipulator systems, it is almost impossible to establish a completely accurate mathematical model, and the existence of modeling errors will inevitably weaken its control effect.

[0006] In addition, some intelligent control methods have also been applied in this field. For example, neural network control has strong nonlinear approximation ability, but it usually requires a large amount of sample data for offline training and has heavy online computation burden. Fuzzy control does not rely on an accurate model, but its rule setting often depends on expert experience, lacking systematic design theory and rigorous stability analysis.

[0007] Although the above control methods have improved the control performance of hydraulic manipulators to some extent, they have not fundamentally solved the core problem of the system, and generally have the following defects:

[0008] (1) Control accuracy and response speed are insufficient: the existing method often shows slow response and low control accuracy when dealing with the high nonlinearity and uncertainty of the system, especially in the working condition requiring fast start-stop or direction change, the dynamic performance is poor, which easily leads to inaccurate positioning of the end effector and too large trajectory tracking error.

[0009] (2) Poor system stability: during the dynamic change of start and brake, the system is prone to excessive oscillation and even instability, affecting the smoothness and safety of the operation.

[0010] (3) Weak anti-disturbance ability: most existing control methods lack the ability to actively perceive and online compensate for external unknown disturbances and internal parameter uncertainties, resulting in poor robustness of the system in actual complex environment, and the error is difficult to effectively suppress.

[0011] In summary, the existing technology still has bottlenecks in control accuracy, response speed and stability when dealing with hydraulic manipulator systems with complex uncertainty. Therefore, it is urgent to develop a new type of control method that can efficiently deal with strong nonlinearity and complex uncertainty to solve the problem of accurate and stable trajectory following of hydraulic manipulator. SUMMARY

[0012] Therefore, the purpose of the present application is to provide a PID-DOB composite driving trajectory control method for hydraulic manipulator, which improves the control accuracy and stability of the hydraulic manipulator.

[0013] To achieve the above purpose, the present application provides the following technical scheme:

[0014] A PID-DOB composite driving trajectory control method for hydraulic manipulator mainly includes the following technical means:

[0015] 1) Subsection control strategy: this scheme divides the trajectory tracking control process of the hydraulic manipulator into three stages: start-up stage, main stage and braking stage. The bias compensation mechanism is used in the start-up stage to overcome the dead zone characteristics of the hydraulic system and improve the effectiveness of the initial control. The PID controller is designed based on the second-order model obtained by system identification in the main stage to realize fine tracking. The three-stage step deceleration strategy is used in the braking stage to suppress overshoot and terminal oscillation and ensure smooth stopping. This subsection control strategy can better adapt to the dynamic response characteristics of the hydraulic system at different stages, significantly improving the tracking stability and control accuracy of the system.

[0016] 2) Compound PID-DOB control method: Based on the traditional PID controller, the disturbance observer structure (DOB) is integrated. The disturbance observer estimates and compensates the external disturbance and modeling error of the system in real time, significantly enhancing the robustness and anti-disturbance ability of the control system under complex working conditions. Through the disturbance observer, the system can actively perceive and compensate for external load disturbance, oil temperature fluctuations and other uncertain factors, ensuring that the hydraulic manipulator maintains a small trajectory error and high stability during actual operation.

[0017] 3) System identification based on experimental data: The experimental data is used to identify the second-order linear model of each control axis of the hydraulic system, guiding the parameter tuning of the PID controller and the design of the DOB observer. This method makes the controller design more targeted, significantly reducing the strong parameter dependence and parameter tuning difficulty in traditional hydraulic control, and improving the universality and easy deployment of the controller. Through system identification, the dynamic model of each execution axis is obtained, a model reference type PID controller is constructed, and the disturbance observer is integrated to compensate for disturbances and unmodeled dynamics in real time. The overall scheme can achieve high-precision trajectory tracking of the hydraulic manipulator end.

[0018] The PID-DOB compound driving trajectory control method specifically includes:

[0019] The hydraulic system of the hydraulic manipulator is modeled, including hydraulic pumps, electro-hydraulic pilot valves, hydraulic cylinders and hydraulic motors, etc. The saturation characteristics and dead zone characteristics of the hydraulic system are analyzed. Then, by analyzing the signal spectrum diagram of the normal operation of the hydraulic manipulator, each axis of the manipulator is identified and modeled to obtain a second-order linear system model of each axis. A segmented PID control method is designed based on the identified second-order linear system model. A compound PID-DOB control rate is designed based on the PID control method, where DOB is a disturbance observer. The disturbance observer estimates and compensates for external disturbances and modeling errors in real time.

[0020] The segmented PID control method divides the control process into start-up segment, main segment and braking segment according to time, speed and position information. In the start-up segment, a bias compensation mechanism is used to overcome the dead zone characteristics of the hydraulic system and improve the effectiveness of the initial control response. In the main segment, a model reference type PID controller is designed for the second-order system model of each axis obtained by system identification to achieve fine tracking. In the braking segment, a three-stage step deceleration strategy is used to suppress overshoot and terminal oscillation and ensure smooth stopping.

[0021] Further, the hydraulic system of the hydraulic manipulator includes hydraulic pumps, electro-hydraulic pilot valves, hydraulic cylinders and hydraulic motors, etc. The hydraulic pump provides hydraulic oil as a power source, the electro-hydraulic pilot valve controls the flow direction and flow rate of the hydraulic oil according to the electrical signal sent by the main control system (such as MATLAB), and the hydraulic cylinder and hydraulic motor drive the movement of the manipulator as execution elements.

[0022] Furthermore, the hydraulic robotic arm exhibits dead-zone characteristics on each axis, meaning that when the input electrical signal is within a certain range, the valve flow is zero, and the robotic arm does not move. The dead-zone range of each axis was determined through experimental testing.

[0023] Turntable: Dead zone upper limit 8940, lower limit 7270. Turn left when the signal is less than 7270, and turn right when the signal is greater than 8940.

[0024] Boom: Dead zone upper limit 8195, lower limit 6990. When the signal is less than 6990, it descends; when it is greater than 8195, it rises.

[0025] Stick: Dead zone upper limit 8400, lower limit 8145. Raise when signal is less than 8145, lower when signal is greater than 8400;

[0026] When the electro-hydraulic valve is fully open, the hydraulic oil flow rate reaches its maximum value. At this point, even if the signal value is further increased or decreased, the flow rate will no longer change, indicating saturation characteristics. The saturation signal values ​​for each axis were determined through step signal experiments.

[0027] Turntable: The speed remains unchanged when the left turn saturation signal is less than 7270, and the speed remains unchanged when the right turn saturation signal is greater than 8940;

[0028] Boom: The speed remains unchanged when the descent saturation signal is less than 6990, and the speed remains unchanged when the lifting saturation signal is greater than 8195;

[0029] Stick control: The speed remains unchanged when the lifting saturation signal is less than 8145, and the speed remains unchanged when the lowering saturation signal is greater than 8400.

[0030] Furthermore, a second-order linear system model for each axis is obtained. Specifically, the system identification method is used to collect input and output data of four actions: boom lifting, boom lowering, stick lifting, and stick lowering, with a step signal as input. The least squares method is then used for parameter estimation.

[0031] Furthermore, the composite PID-DOB control rate design includes DOB design, segmented trajectory tracking control rate design, start-up control rate design based on bias compensation and PID control, main body control rate design based on a second-order system identification model, and three-stage braking control rate design.

[0032] Furthermore, the DOB design specifically includes: employing an estimated value of the interference signal d. Then connect a low-pass filter Q(s); and then connect the actual system's transfer function G. p (s) replaced with the nominal model G n The inverse G of (s) n -1 (s); The input signal c of the PID control is added to the interference signal d, and then the interference is subtracted to obtain the compensation signal by the interference observer.

[0033]

[0034] Where u is the output of the control law.

[0035] Furthermore, the segmented trajectory tracking control design specifically includes: after obtaining the reference joint angle using the composite PID-DOB control, the error is compared with the real-time joint angle obtained by the sensor. Based on the independent variable time, it is determined whether the current control is in the start-up phase, main phase, or braking phase, thereby activating different control strategies. Then, the composite controller inputs the electrical control signal to the electro-hydraulic pilot valve, i.e., the drive layer. The controller needs to overcome the nonlinear characteristics of hydraulics: dead zone characteristics, saturation characteristics, and time delay characteristics. The electro-hydraulic pilot valve converts the electrical signal into the flow direction and velocity of the hydraulic oil, which then flows to the hydraulic cylinder to drive the hydraulic robotic arm to complete the corresponding action. The signals transmitted alternately in each process are converted from angle to electrical signal, then to hydraulic robotic arm rotation, and finally returned by the sensor.

[0036] Furthermore, the design of the start-up control rate based on bias compensation and PID control specifically includes: setting the first period of the entire trajectory as the start-up segment; the feedback closed-loop strategy of the start-up segment, i.e., the model-free PID control method, obtains the reference point of the trajectory planning stage through the independent variable time t, i.e., the reference position Q of each axis of the unmanned hydraulic robotic arm. r (θ 1r ,θ 2r ,θ 3r The target value is used as the initial value; then the tilt angle q is obtained by a tilt sensor mounted on the hydraulic robotic arm. s (θ 1s ,θ 2s ,θ 3s Based on the relative positional relationship between the sensor and the hydraulic robotic arm, angle compensation conversion is performed to obtain the real-time angles q(θ1,θ2,θ3) of each axis of the unmanned hydraulic robotic arm, which are then used as measured values. The deviation, the change in deviation, and the cumulative deviation are calculated by subtracting the measured value from the target value. The above results are then processed through the proportional, integral, and derivative elements of a PID control system to calculate the control signal u. Finally, the control signal is input to the hydraulic drive system, which acts on the electro-hydraulic pilot valve to complete the rotation of the corresponding hydraulic robotic arm. Because the hydraulic unmanned steel grabber has dead zone nonlinear characteristics, the planned trajectory cannot be quickly and stably tracked in the startup phase using only a feedback closed-loop strategy. Therefore, this invention uses offset compensation to overcome the dead zone. The offset compensation value is referenced to the upper limit of the dead zone under the corresponding action of each axis.

[0037] Further, the subject section control rate design based on the second-order system identification model specifically includes: the subject section and the starting section are similar to the PID control based on the fact that the models of the four actions of boom lifting, boom lowering, stick lifting and stick lowering identified by the least square method are zero points of the upper and lower limits of the dead zone, so the designed PID control rate also needs to add the dead zone compensation. The expression of the PID control rate of the boom is:

[0038]

[0039] Wherein, u[k] is the control signal at the current time (the signal value input to the electro-hydraulic proportional valve), e[k] is the control error at the current time, that is, the difference between the reference angle and the actual angle, Δu1 is the dead zone compensation value when moving forward, Δu2 is the dead zone compensation value when moving backward, K pu , K iu , K du are the PID parameters when rising, K pd , K id , K dd are the PID parameters when falling; because the deviation used in the application is the difference between the radians of the mechanical arm, the value is relatively small, and the corresponding signal value is 2000-11000, in order to facilitate debugging and avoid K p being too large, the application multiplies the deviation by 100 as the proportional term. Similarly, the expression of the PID control rate of the stick is:

[0040]

[0041] Because the lifting signal of the boom is below the zero point, and the lowering signal is above the zero point; and the stick is just the opposite, the lifting signal of the stick is above the zero point, and the lowering signal is below the zero point, so the expression of the PID control rate of the boom is opposite to that of the stick;

[0042] At this time, the PID part of the subject section has been designed, and the disturbance observer (DOB) needs to be added to the PID controller, and the difference between the nominal model and the real model identified by the feedforward compensation is calculated, that is, the output u3(k) of the DOB is calculated, and then the noise is removed through the filter to obtain the feedforward input u o (k);

[0043]

[0044] Wherein, G n(z) is a discrete transfer function of nominal model, z is a complex variable of Z transform, num1, num2, num3 are coefficients of a discrete nominal model numerator, den1, den2, den3 are coefficients of a discrete nominal model denominator; y(k) is an output signal (joint angle) of a single joint of the hydraulic arm at a current time, d2, d3, d4 are discrete coefficients of a denominator polynomial, n1, n2, n3, n4 are discrete coefficients of a numerator polynomial;

[0045] The main part of the composite PID-DOB control rate algorithm needs to input the current axis PID parameter, the initialization flag, the reference angle of the axis at the current time, the real-time angle of the axis at the current time, the direction flag of the axis at the current time, the sampling time and the end point of the reference trajectory, and outputs the control signal of the axis at the current time.

[0046] Further, the three-section brake section control rate design specifically includes: in order to balance the stability and rapidity of braking, the braking section adopts a three-section processing strategy, Q dis = |Q real -Q refEnd |, wherein Q dis is the distance of the real-time joint position from the end point of the reference trajectory, Q real is the real value of the real-time joint position, and Q refEnd is the reference value of the real-time joint position; taking the boom lowering as an example, the three sections are: 1) when the proportional term 100*e[k] is less than or equal to -2 and Q dis is greater than 0.01; 2) when the proportional term 100*e[k] is greater than -2 and Q dis is greater than 0.01; and 3) when Q dis is less than or equal to 0.01. The first section of the braking section has a relatively large deviation, the absolute value of the proportional term is greater than 2 radians, and the distance from the end point is also greater than 0.01 radian at this time, so the deviation is large and there is still a large distance from the end point to stop, so the strategy adopted by the first section of the present application is still a feedback closed-loop strategy, i.e., a PID control method is used to control the trajectory tracking of the first section; the second section of the braking section has a relatively small deviation, the absolute value of the proportional term is less than 2 radians, and the distance from the end point Q dis is also relatively large, i.e., greater than 0.01 radian at this time, so a small brake step signal is used, which can effectively slow down the final stop without violent shaking, and also prevent the hydraulic arm from suddenly stopping too far from the end point; the third section of the braking has a small distance from the end point, i.e., Q dis is less than 0.01 radian, so a zero signal of 8190 is directly given to make the hydraulic arm stop immediately, and the zero signal of the third section is not very abrupt due to the deceleration of the second section, so the hydraulic arm can be quickly stopped without much shaking.

[0047] The beneficial effects of this invention are as follows:

[0048] 1) High-precision trajectory tracking: By employing a segmented control strategy, combined with a disturbance observer (DOB) and system identification methods, this scheme significantly improves the trajectory tracking accuracy of the hydraulic robotic arm. Specifically, the starting stage uses an offset compensation mechanism to overcome the dead zone characteristics of the hydraulic system, the main stage uses a PID controller designed based on a second-order model, and the braking stage adopts a three-stage stepped deceleration strategy. These measures work together to enable the hydraulic robotic arm to achieve high-precision trajectory tracking under complex working conditions.

[0049] 2) Enhanced robustness and disturbance rejection: This scheme introduces a disturbance observer (DOB) to estimate and compensate for external disturbances and modeling errors in real time. This method significantly enhances the robustness and disturbance rejection of the control system under complex operating conditions, ensuring that the hydraulic robotic arm maintains small trajectory errors and high stability in actual operation. By actively suppressing uncertainties such as external load disturbances and oil temperature fluctuations through the disturbance observer, the performance of the control system is significantly improved.

[0050] 3) Simplified parameter tuning process: Second-order linear model identification of each control axis of the hydraulic system is performed using experimental data to guide PID controller parameter tuning and DOB observer design. This method makes controller design more targeted, significantly reducing the problems of strong parameter dependence and difficult parameter tuning in traditional hydraulic control, and improving the controller's versatility and ease of deployment. By obtaining the dynamic model of each execution axis through system identification, a model reference PID controller is constructed, and a disturbance observer is integrated to compensate for disturbances and unmodeled dynamics in real time. The overall solution can achieve high-precision trajectory tracking at the end effector of the hydraulic robotic arm.

[0051] In summary, this solution has significant technical advantages, not only improving the control accuracy and stability of the hydraulic robotic arm, but also enhancing its robustness and anti-disturbance ability under complex working conditions. At the same time, it simplifies the parameter adjustment process and improves engineering applicability and transfer efficiency.

[0052] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0053] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0054] Figure 1Flow chart of hydraulic manipulator compound PID-DOB trajectory tracking control method for scrap steel grabbing scene;

[0055] Figure 2 Schematic diagram for modeling and analyzing hydraulic system;

[0056] Figure 3 Schematic diagram for analyzing nonlinear system of hydraulic system;

[0057] Figure 4 Schematic diagram for designing compound controller;

[0058] Figure 5 System block diagram of segmented PID control;

[0059] Figure 6 Flow chart of segmented PID control method;

[0060] Figure 7 Flow chart of compound PID-DOB control rate design;

[0061] Figure 8 System block diagram of segmented trajectory tracking control design. DETAILED DESCRIPTION

[0062] The present application will be described in detail below with specific reference being made to certain embodiments thereof. The advantages and effects of the present application can be easily understood by those skilled in the art from the contents disclosed in the present specification. The present application can also be implemented or applied in other different embodiments, and the details in the present specification can be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the drawings provided in the following embodiments only schematically illustrate the basic concept of the present application, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0063] Please refer to Figures 1-8 The embodiment of the present application provides a hydraulic manipulator compound PID-DOB trajectory tracking control method for scrap steel grabbing scene. First, the overall analysis of the hydraulic driving system is performed, and then the dead zone characteristics and saturation characteristics of the hydraulic driving are analyzed. Then, the signal spectrum diagram of the hydraulic scrap steel grabbing machine is analyzed to identify and model each axis of the manipulator, and a segmented PID control method is designed based on the identified second-order system model. A disturbance observer is added for improvement, and a compound PID-DOB control rate is designed.

[0064] 1. Hydraulic manipulator system analysis

[0065] As Figure 2As shown, the hydraulic system of the hydraulic steel grabber mainly consists of hydraulic pump, electro-hydraulic pilot valve, hydraulic cylinder and hydraulic motor and other elements. The hydraulic pump provides hydraulic oil as the power source, the electro-hydraulic pilot valve controls the flow direction and flow rate of the hydraulic oil according to the electrical signal sent by the main control system (such as MATLAB), and the hydraulic cylinder and hydraulic motor are used as the execution element to drive the mechanical arm movement. The hydraulic system adopts a series circuit design, and the three hydraulic cylinders are supplied by the same hydraulic pump, which leads to a slower hydraulic flow rate of a certain axis during multi-axis linkage than during single-axis movement, increasing the control difficulty. The working principle of the electro-hydraulic valve is to control the valve direction and opening degree according to the difference between the input electrical signal and the zero signal: the difference determines the flow direction, and the difference size is proportional to the flow rate.

[0066] As shown in Figure 3 , each axis of the hydraulic steel grabber has a dead zone characteristic, that is, when the input electrical signal is within a certain range, the valve port flow is zero, and the mechanical arm has no action. Through experimental testing, the dead zone range of each axis is obtained:

[0067] Rotary table: upper limit of dead zone 8940, lower limit 7270, signal less than 7270, left turn, greater than 8940, right turn.

[0068] Boom: upper limit of dead zone 8195, lower limit 6990, signal less than 6990, lower, greater than 8195, lift.

[0069] Dipper arm: upper limit of dead zone 8400, lower limit 8145, signal less than 8145, lift, greater than 8400, lower.

[0070] When the electro-hydraulic valve is fully opened, the hydraulic oil flow rate reaches the maximum value, at which time even if the signal value continues to increase or decrease, the flow rate will no longer change, that is, saturation characteristics appear. Through step signal experiment, the saturation signal value of each axis is determined:

[0071] Rotary table: left turn saturation signal less than 7270, speed unchanged, right turn saturation signal greater than 8940, speed unchanged.

[0072] Boom: lower saturation signal less than 6990, speed unchanged, lift saturation signal greater than 8195, speed unchanged.

[0073] Dipper arm: lift saturation signal less than 8145, speed unchanged, lower saturation signal greater than 8400, speed unchanged.

[0074] Due to the difficulty in identifying nonlinear models of hydraulic systems, a second-order linear system model is used to approximate the nonlinear hydraulic system. Using a system identification method, a step signal is used as input to collect input and output data for four actions: boom raising, boom lowering, stick raising, and stick lowering. Parameter estimation is performed using the least squares method. In this embodiment, four second-order models are established for each of the four actions: boom raising, boom lowering, stick raising, and stick lowering, thus performing system identification for each of these four actions.

[0075] Table 1 Four Second-Order Models

[0076]

[0077] In Table 1, s represents the complex frequency domain variable in the Laplace transform, and e represents the exponent symbol in scientific notation.

[0078] 2. Composite PID-DOB control law design (e.g.) Figure 4 (As shown)

[0079] 1) Disturbance Observer (DOB) Design:

[0080] Using the estimated value of the interference signal d A low-pass filter Q(s) is then connected. The transfer function G of the actual system is then applied. p (s) replaced with the nominal model G n The inverse G of (s) n -1 (s). The final framework is as follows: Figure 7 As shown, the area within the dashed box is the interference observer. From Figure 7 As can be seen from this, u is the output of the control law, which is the input signal c of the PID control plus the disturbance signal d, and then the disturbance is subtracted to obtain the compensation signal by the observer.

[0081]

[0082] 2) Segmented trajectory tracking control law design:

[0083] Because the track is a trajectory with time information, this embodiment divides the entire trajectory into three segments: the starting segment, the main segment, and the braking segment, based on time, speed, and position information.

[0084] Start-up phase: An offset compensation mechanism is used to overcome the dead zone of the hydraulic system and improve the effectiveness of control in the initial response phase;

[0085] Main body: Based on the second-order transfer function models of each axis identified by the system, a model reference PID controller is designed to achieve fine-grained tracking;

[0086] Braking section: three-stage step deceleration strategy is adopted to suppress overshoot and terminal oscillation, and to ensure smooth stopping.

[0087] The reference joint angle obtained by the composite PID-DOB control rate is compared with the real-time joint angle obtained by the sensor to obtain the error. According to the independent variable time, it is judged whether the current is in the control rate starting section, the main section or the braking section, so as to start different control strategies. Then the composite controller inputs the electric control signal to the electro-hydraulic proportional valve, that is, the driving layer. The controller needs to overcome the nonlinear characteristics of the hydraulic pressure: dead zone characteristics, saturation characteristics and time delay characteristics. The electro-hydraulic proportional valve converts the electric signal into the flow direction and flow rate of the hydraulic oil, and the hydraulic oil will flow to the hydraulic cylinder to drive the hydraulic mechanical arm to complete the corresponding action, which is the whole process from trajectory planning to execution layer. The whole loop can be represented as a process Figure 7 , and the signals alternately transmitted by each process are converted from angle to electric signal and then to hydraulic mechanical arm rotation and finally returned by the sensor.

[0088] 3) Starting section control rate design based on bias compensation and PID control:

[0089] The first section of the whole trajectory is set as the starting section. The feedback closed-loop strategy of the starting section is the model-free PID control method. The reference point in the trajectory planning stage is obtained by the independent variable time t, that is, the reference position of each axis of the unmanned hydraulic steel grabbing machine, which is taken as the target value. Then the measured inclination is obtained by the inclination sensor installed on the hydraulic mechanical arm, and the angle is compensated and converted according to the relative position relationship between the sensor and the hydraulic mechanical arm, to obtain the real-time angle of each axis of the unmanned hydraulic steel grabbing machine and take it as the measured value. The deviation, the change of the deviation and the accumulation of the deviation are calculated by subtracting the measured value from the target value. The above results are calculated through the proportional, integral and differential links of the PID control, so as to calculate the control signal u. Finally, the control signal is input to the hydraulic drive system to act on the electro-hydraulic valve to complete the rotation of the corresponding hydraulic mechanical arm. Because the hydraulic unmanned steel grabbing machine has the dead zone nonlinear characteristic, it is impossible to quickly and stably track the planned trajectory in the starting section only by the feedback closed-loop strategy, so the bias compensation is adopted to overcome the dead zone. The bias compensation value refers to the upper limit value of the dead zone under the corresponding action of each axis.

[0090] 4) Main section control rate design based on second-order system identification model:

[0091] The main section is also similar to the starting section and is based on PID control. Because the models of the boom lifting, boom lowering, stick lifting and stick lowering four actions identified by the least square method in the above are taken as the zero point of the dead zone upper and lower limits, the dead zone compensation needs to be added when designing the PID control rate. The PID control rate expression of the boom is:

[0092]

[0093] wherein, K pu , K iu , K du are PID parameters when rising, K pd , K id , K dd are PID parameters when falling. Because the deviation adopted in this embodiment is the difference of the radian of the mechanical arm, its value is relatively small, and the corresponding signal value is 2000-11000, in order to facilitate debugging and avoid K p being too large, this embodiment multiplies the deviation by 100 as the proportional term. For the same reason, the PID control rate expression of the dipper arm is:

[0094]

[0095] Because the lifting signal of the boom is below the zero point, and the falling signal is above the zero point, and the dipper arm is just the opposite, its lifting signal is above the zero point, and the falling signal is below the zero point, so its PID control rate expression is just the opposite of the PID control rate expression of the boom.

[0096] At this time, the PID part of the main section has been designed, and the disturbance observer needs to be added to the PID controller, and the difference between the identified nominal model and the real model is compensated through feedforward, as shown in Figure 7 , the u(k) of formula (3) is c(k) in Figure 7 , the nominal model is a second-order system, and after discretization, it is like formula (4), wherein num1, num2, num3 are the coefficients of the numerator, and den1, den2, den3 are the coefficients of the denominator. According to Figure 7 , the output of the disturbance observer can be calculated, that is, formula (5), and finally the noise is removed through the filter to obtain the feedforward input u o (k).

[0097]

[0098] The main part of the composite PID-DOB control rate algorithm needs to input the current axis PID parameter, the initialization flag, the reference angle of the current axis at the current time, the real-time angle of the current axis at the current time, the direction flag of the current axis at the current time, the sampling time, and the end point of the reference trajectory, and output the control signal of the current axis at the current time.

[0099] 5) Three-section brake section control rate design:

[0100] In order to take into account the stability and rapidity of braking, the strategy adopted in this embodiment is divided into three sections, that is, Q dis = |Q real -Q refEnd |, that is, Q disThe distance of the real-time joint position to the end point of the reference trajectory is taken as an example of the bucket rod descending, and the three segments are respectively:

[0101] (1) When the proportional term 100*e[k] is less than or equal to -2 and Q dis > 0.01;

[0102] (2) When the proportional term 100*e[k] is greater than -2 and Q dis > 0.01;

[0103] (3) When Q dis ≤ 0.01.

[0104] Firstly, the first segment of the braking segment has a relatively large deviation, and the absolute value of the proportional term is greater than 2 radians, and at this time, the distance to the end point is also greater than 0.01 radians, so it can be seen that the deviation is large and there is still a large distance to stop at the end point, so the strategy adopted in the first segment of the embodiment is still a feedback closed-loop strategy, that is, a PID control method is used to control the trajectory tracking of the first segment. The second segment of the braking segment means that the deviation is relatively small at this time, and the absolute value of the proportional term is less than 2 radians; but at this time, the distance to the end point is still relatively large, that is, greater than 0.01 radians, at this time, a small brake step signal is adopted, on the one hand, it can effectively slow down so that the final stop will not be violent, on the other hand, it will not cause the hydraulic manipulator to suddenly stop too far from the end point. The third segment of the braking segment means that the distance to the end point is small, that is, Q dis is less than 0.01 radians, at this time, a zero signal of 8190 is directly given to make the hydraulic manipulator stop immediately, and at this time, due to the deceleration of the second segment of the brake, the zero signal of the third segment will not be too abrupt, and the hydraulic manipulator can be quickly stopped without much shaking.

[0105] Embodiment 1:

[0106] The effectiveness of the PID control method is verified on a 10-ton small hydraulic steel grabbing machine and a 39-ton medium hydraulic steel grabbing machine, and a comparative experiment of two control rates is carried out on the small hydraulic steel grabbing machine, and the results show that the average dynamic error of the rising compound action compound control method is reduced to 74.17% compared with the PID, and the falling compound action is reduced to 61.92%, which proves that the dynamic tracking accuracy of the compound PID-DOB control rate is higher.

[0107] In summary, the trajectory tracking control process of the hydraulic manipulator is innovatively divided into three stages of starting segment, main segment and braking segment, different control rate structures are designed according to the dynamic response characteristics of the hydraulic system in different stages, the overall control process is more accurate and efficient, and the tracking stability and control accuracy of the system are significantly improved.

[0108] Based on the traditional PID controller, the disturbance observer structure (DOB) is fused to estimate and compensate the external disturbance and modeling error in real time, which significantly enhances the robustness and anti-disturbance ability of the control system under complex working conditions, and solves the performance degradation problem of the hydraulic system caused by load fluctuation and nonlinear factors.

[0109] The second-order linear model identification of each control axis of the hydraulic system is performed by using experimental data to guide the PID controller parameter tuning and DOB observer design, so that the controller design is more targeted, the strong parameter dependence and difficult parameter tuning problem in the traditional hydraulic control are significantly reduced, and the universality and easy deployment of the controller are improved.

[0110] Finally, it should be pointed out that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the purpose and scope of the technical solutions, which should be covered in the scope of the claims of the present application.

Claims

1. A PID-DOB composite drive trajectory control method for hydraulic robotic arms, characterized in that, The method includes: The hydraulic system of the hydraulic robotic arm is modeled, and its saturation and dead-zone characteristics are analyzed. Then, by analyzing the signal spectrum of the robotic arm during normal operation, each axis of the robotic arm is identified and modeled, resulting in a second-order linear system model for each axis. Based on the identified second-order linear system models, a piecewise PID control method is designed. A composite PID-DOB control law is designed based on the PID control method, where DOB is a disturbance observer. DOB estimates and compensates for external disturbances and modeling errors in real time. The segmented PID control method divides the control process into a starting segment, a main segment, and a braking segment based on time, speed, and position information. In the starting segment, an offset compensation mechanism is used to overcome the dead zone characteristics of the hydraulic system. In the main segment, a model reference PID controller is designed for each axis based on the second-order system model identified by the system. In the braking segment, a three-stage stepped deceleration strategy is adopted to suppress overshoot and terminal oscillation.

2. The PID-DOB composite drive trajectory control method according to claim 1, characterized in that, The hydraulic system of the hydraulic robotic arm includes a hydraulic pump, an electro-hydraulic pilot valve, a hydraulic cylinder, and a hydraulic motor. The hydraulic pump serves as a power source to supply hydraulic oil, the electro-hydraulic pilot valve controls the flow direction and speed of the hydraulic oil according to the electrical signal sent by the main control system, and the hydraulic cylinder and hydraulic motor serve as actuators to drive the movement of the robotic arm.

3. The PID-DOB composite drive trajectory control method according to claim 1, characterized in that, Each axis of the hydraulic robotic arm has a dead zone characteristic, that is, when the input electrical signal is within a certain range, the flow rate at the valve port is zero and the robotic arm does not move; the dead zone range of each axis was obtained through experimental testing. When the electro-hydraulic valve is fully opened, the hydraulic oil flow rate reaches its maximum value. At this point, even if the signal value is increased or decreased, the flow rate will no longer change, i.e., saturation characteristics appear. The saturation signal value of each axis is determined by step signal experiment.

4. The PID-DOB composite drive trajectory control method according to claim 1, characterized in that, The second-order linear system model of each axis is obtained. Specifically, the system identification method is used to collect the input and output data of four actions: boom lifting, boom lowering, stick lifting, and stick lowering, with the step signal as input. The least squares method is used for parameter estimation.

5. The PID-DOB composite drive trajectory control method according to claim 1, characterized in that, The composite PID-DOB control rate design includes DOB design, segmented trajectory tracking control rate design, start-up control rate design based on bias compensation and PID control, main body control rate design based on a second-order system identification model, and three-stage braking control rate design.

6. The PID-DOB composite drive trajectory control method according to claim 5, characterized in that, The DOB design specifically includes: using an estimated value of the interference signal d. Then connect a low-pass filter Q(s); and then connect the actual system's transfer function G. p (s) replaced with the nominal model G n The inverse G of (s) n -1 (s); The input signal c of the PID control is added to the interference signal d, and then the interference is subtracted to obtain the compensation signal by the interference observer. : Where u is the output of the control law.

7. The PID-DOB composite drive trajectory control method according to claim 5, characterized in that, The segmented trajectory tracking control design specifically includes: after obtaining the reference joint angle using a composite PID-DOB control, the error is compared with the real-time joint angle obtained by the sensor. Based on the independent variable time, it is determined whether the current control is in the start-up phase, main phase, or braking phase, thereby activating different control strategies. Then, the composite controller inputs the electrical control signal to the electro-hydraulic pilot valve, i.e., the drive layer. The controller needs to overcome the nonlinear characteristics of hydraulics: dead zone characteristics, saturation characteristics, and time delay characteristics. The electro-hydraulic pilot valve converts the electrical signal into the flow direction and velocity of the hydraulic oil, which then flows to the hydraulic cylinder to drive the hydraulic robotic arm to complete the corresponding action. The signals transmitted alternately in each process are converted from angle to electrical signal, then to hydraulic robotic arm rotation, and finally returned by the sensor.

8. The PID-DOB composite drive trajectory control method according to claim 5, characterized in that, The design of the start-up control rate based on bias compensation and PID control specifically includes: setting the first period of the entire trajectory as the start-up segment; the feedback closed-loop strategy of the start-up segment, i.e., the model-free PID control method, obtains the reference point of the trajectory planning stage through the independent variable time t, i.e., the reference position Q of each axis of the unmanned hydraulic robotic arm. r (θ 1r ,θ 2r ,θ 3r The target value is used as the initial value; then the tilt angle q is obtained by a tilt sensor mounted on the hydraulic robotic arm. s (θ 1s ,θ 2s ,θ 3s Based on the relative positional relationship between the sensor and the hydraulic robotic arm, angle compensation conversion is performed to obtain the real-time angles q(θ1,θ2,θ3) of each axis of the unmanned hydraulic robotic arm, which are then used as measured values. The deviation, the change in deviation, and the cumulative deviation are calculated by subtracting the measured value from the target value. The above results are then processed through the proportional, integral, and derivative elements of a PID control system to calculate the control signal u. Finally, the control signal is input to the hydraulic drive system, which acts on the electro-hydraulic pilot valve to complete the rotation of the corresponding hydraulic robotic arm. Offset compensation is used to overcome the dead zone. The offset compensation value is referenced to the upper limit of the dead zone under the corresponding action of each axis.

9. The PID-DOB composite drive trajectory control method according to claim 5, characterized in that, The main body control rate design based on the second-order system identification model specifically includes: Similar to the starting section, the main body control is based on PID control, and the PID control rate expression for the boom is: Where u[k] is the control signal at the current moment, i.e., the signal value input to the electro-hydraulic proportional valve; e[k] is the control error at the current moment, i.e., the difference between the reference angle and the actual angle; Δu1 is the dead zone compensation value during forward motion; Δu2 is the dead zone compensation value during reverse motion, K pu K iu K du K is the PID parameter during the rise. pd K id K dd The PID parameters are for the descent phase; similarly, the PID control rate expression for the stick is: The PID control rate expression for the boom is the opposite of the PID control rate expression for the stick. At this point, the PID control section of the main body is complete. The DOB needs to be added to the PID controller. Then, the difference between the nominal and true models is identified through feedforward compensation, i.e., the output u3(k) of the DOB is calculated. Finally, noise is removed using a filter to obtain the feedforward input u. o (k); u o (k)=-d2u o (k-1)-d3u o (k-2)-d4u o (k-3)+n1u3(k)+n2u3(k-1)+n3u3(k-2)+n4u3(k-3) Among them, G n (z) is the discrete transfer function of the nominal model, z is the complex variable of the Z-transform, num1, num2, num3 are the coefficients of the numerator of the discrete nominal model, den1, den2, den3 are the coefficients of the denominator of the discrete nominal model; y(k) is the output signal of a single joint of the hydraulic arm at the current moment, i.e., the joint angle; d2, d3, d4 are the discrete coefficients of the denominator polynomial, and n1, n2, n3, n4 are the discrete coefficients of the numerator polynomial; The main part of the composite PID-DOB control algorithm requires the PID parameters of the current axis, the initialization flag, the reference angle of the axis at the current moment, the real-time angle of the axis at the current moment, the direction flag of the axis at the current moment, the sampling time, and the endpoint of the reference trajectory as inputs. The output is the control signal of the axis at the current moment.

10. The PID-DOB composite drive trajectory control method according to claim 5, characterized in that, The three-stage braking control rate design specifically includes: the braking strategy is divided into three stages, letting Q... dis =|Q real -Q refEnd |, where Q dis This refers to the real-time distance between the joint position and the endpoint of the reference trajectory. In the first braking segment, the absolute value of the deviation proportional term is greater than 2 radians, and the distance to the endpoint is greater than 0.01 radians. A feedback closed-loop strategy, i.e., PID control, is used to control the trajectory tracking in the first segment. In the second braking segment, the absolute value of the deviation proportional term is less than 2 radians, and the distance to the endpoint is Q... dis When the angle is greater than 0.01 radians, a braking step signal is used; the third braking segment is the distance Q from the endpoint. dis If the angle is less than 0.01 radians, a zero-point signal is given to immediately stop the hydraulic robotic arm.

Images