CN113341943B - Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning - Google Patents
Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning Download PDFInfo
- Publication number
- CN113341943B CN113341943B CN202010140701.0A CN202010140701A CN113341943B CN 113341943 B CN113341943 B CN 113341943B CN 202010140701 A CN202010140701 A CN 202010140701A CN 113341943 B CN113341943 B CN 113341943B
- Authority
- CN
- China
- Prior art keywords
- disturbance
- course angle
- angle
- distance error
- control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 35
- 238000013459 approach Methods 0.000 claims abstract description 4
- 238000000034 method Methods 0.000 claims description 34
- 230000008569 process Effects 0.000 claims description 13
- 238000012821 model calculation Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 5
- 238000005259 measurement Methods 0.000 claims description 4
- 230000036039 immunity Effects 0.000 claims 1
- 238000003801 milling Methods 0.000 claims 1
- 230000008859 change Effects 0.000 abstract description 5
- 230000000694 effects Effects 0.000 description 5
- 230000001629 suppression Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 3
- 241000234314 Zingiber Species 0.000 description 2
- 235000006886 Zingiber officinale Nutrition 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005034 decoration Methods 0.000 description 2
- 235000008397 ginger Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000011435 rock Substances 0.000 description 2
- 238000005096 rolling process Methods 0.000 description 2
- 241000408659 Darpa Species 0.000 description 1
- 241000907661 Pieris rapae Species 0.000 description 1
- 239000012190 activator Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 210000001503 joint Anatomy 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000002787 reinforcement Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0225—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving docking at a fixed facility, e.g. base station or loading bay
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0276—Control of position or course in two dimensions specially adapted to land vehicles using signals provided by a source external to the vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning, which comprises the following steps: step 1, calculating an actual distance error of the unmanned vehicle according to a target track of the unmanned vehicle; step 2, calculating a target course angle of the unmanned vehicle by an anti-interference controller of the distance error, so that the track tracking distance error of the unmanned vehicle tends to zero; step 3, calculating a target steering wheel corner or a hydraulic system corner of the unmanned vehicle by the anti-interference controller of the heading angle, so that the actual heading angle of the unmanned vehicle approaches to the target heading angle obtained in the step 2; and 4, sending the target steering wheel angle or the hydraulic system angle obtained in the step 3 to a steering wheel angle or hydraulic system angle controller for closed-loop control. The invention can obviously improve the adaptability of the algorithm to uncertainty such as vehicle state change, road condition change and the like.
Description
Technical Field
The invention relates to the technical field of unmanned driving, in particular to a repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning.
Background
Accurate tracking (hereinafter, referred to as tracking) of a target traveling track by an unmanned vehicle is one of the most important functions thereof. Proportional-integral-derivative (PID) control is the most widely used tracking control algorithm, and documents (1)
The system "PID Controllers for Autonomous Vehicle Path following" (2017) introduced the use of PID to solve Vehicle tracking control. Since the state of the vehicle changes over time and the road conditions have uncertainty, it is difficult for the PID controller to maintain the best quality of control under different conditions. Although the PID control of the parameter segment can improve the performance to some extent, the segment adjustment of the parameter is complicated, which affects the development efficiency.
Control algorithms based on vehicle and road geometry are also a class of solutions that are more widely used. The pure tracking (pure pursuit) algorithm was First described in the literature (R.Wallace, A. Stentz, C.E.Thorpe, H.Maravec, W.Whittaker, and T.Kanade, "First results in robot road-following", in IJCAI, pp.1089-1095, 1985.). The method cannot give a solution when the distance between the vehicle and the target track is too far, and more importantly, the method can easily induce the oscillation of the vehicle running track when the vehicle runs at a high speed. Another geometry-based control algorithm is the rear wheel based feedback control (rear wheel position based feedback), the literature (C.Samson, "Path following and time-varying feedback stabilization of a wheel mobile robot," in 2nd Int. Conf. On Automation, robotics and Computer Vision, 1992.) introduces the detailed principles and implementations of the algorithm. Similarly, front wheel based feedback control is also described in detail in (front wheel position based feedback) in (M.D. Sensors, "Stanley: the robot where The DARPA Grand Challenge," Journal of field Robotics, vol.23, pp.661-692, 2006). The three methods described above have practical application in DARPAR games held in the united states around 2004. However, because of the need for more accurate vehicle and road geometry information, the performance in tracking accuracy is limited based on geometry and control algorithms, and is mainly applicable to low vehicle speed conditions. Performance degradation is inevitable under different road conditions and vehicle conditions due to the lack of a learning mechanism.
The tracing control algorithm based on the kinematic model is also an algorithm widely adopted in the field of unmanned driving. An article (Y. Kanayama, Y. Kimura, F. Miyazaki, and T. Noguchi, "A stable tracking control method for an autonomous mobile robot," in International Conference on Robotics and Automation, pp.384-389, IEEE, 1990.) describes a vehicle tracking control algorithm based on the Lyapunov function. However, the Lyapunov function can only ensure stability, and the actual trajectory tracking effect needs to be adjusted in detail. To accommodate operation at higher vehicle speeds, methods of output feedback linearization have been proposed by The scholars in The literature (B.d' Andrewa Novel, G.Campo, and G.Basin, "Control of non-polar mobile clients by state feedback linearization," The International journal of particulate research, vol.14, pp.543-559, 1995.). However, this method also generally requires a relatively accurate kinematic model of the vehicle.
The above method is generally applicable to conventional driving situations. For tracking control in case of drift or emergency, model Predictive Control (MPC) is considered as a relatively effective solution. The literature (p.falcone, f.borrelli, j.asgari, h.e.tseng, and d.htovat, "Predictive active steering Control for autonomous vehicle Systems," Transactions on Control Systems Technology, vol.15, pp.566-580, 2007.) compares earlier the application of MPC algorithms to vehicle tracking Control. A series of derivation algorithms are proposed in succession, such as (g.v. raffo, g.k.gos, j.e. normal-Rico, c.r.kelber, and l.b. becker, "a predictive controller for an autonomous vehicle path tracking," Transactions on Intelligent transfer Systems, vol.10, pp.92-102, 2009.) and (y.yoon, j.shin, h.j.kim, y.park, and s.savery, "Model-predictive active and inactive activity for automatic ground vehicles," Control Engineering Practice, vol.17, pp.741-750, 2009 "), and (E.Kim, J.Kim, and M.Sunwoo," Model predictive Control protocol for smooth path tracking of automatic vehicles with activator dynamics, "International Journal of automatic Technology, vol.15, pp.1155-1164, 2014.). However, the MPC method also has a high requirement on the accuracy of the model, and uncertainty caused by model misalignment often causes deterioration of the tracking effect. In addition, MPC requires a high computational cost, and is limited in application to a less computationally intensive platform such as an embedded system.
In addition to the above algorithm, a Linear Parameters Varying (LPV) model is a control algorithm which has been widely used recently. The algorithm adopts an expression form of a linear model, and the accuracy of the nonlinear model is approximated by the change of model parameters along with the working condition. The algorithm is used in the literature (P.G.sp.r, Z.Szab Lou, and J.Bokor, "LPV design of fault-tolerant control for road vehicles," International Journal of Applied Mathematics and Computer Science, vol.22, pp.173-182, 2012). In essence, the LPV uses a model parameter segmentation method to improve the accuracy of a linear model and reduce the influence of uncertainty factors. But this also results in a high complexity of the algorithm, which affects its practicality.
Disclosure of Invention
The invention aims to provide a repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning aiming at the technical defects in the prior art.
The technical scheme adopted for realizing the purpose of the invention is as follows:
a repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning comprises the following steps:
step 1, calculating an actual distance error of the unmanned vehicle according to a target track of the unmanned vehicle;
step 2, calculating a target course angle of the unmanned vehicle by the anti-interference controller of the distance error through the actual distance error obtained in the step 1, the instant disturbance of the distance error obtained by the extended state observer and the accumulated disturbance of the distance error obtained by the iterative learner of the instant total disturbance of the distance error, so that the track tracking distance error of the unmanned vehicle tends to zero;
step 3, the disturbance rejection controller of the heading angle calculates a target steering wheel corner or a hydraulic system corner of the unmanned vehicle through the target heading angle obtained in the step 2, the instant disturbance of the heading obtained by the extended state observer and the accumulated disturbance of the heading angle obtained by the iterative learner of the instant total disturbance of the heading angle, so that the actual heading angle of the unmanned vehicle approaches to the target heading angle obtained in the step 2;
and 4, sending the target steering wheel angle or the hydraulic system angle obtained in the step 3 to a steering wheel angle or hydraulic system angle controller, and performing closed-loop control on the actual steering wheel angle or hydraulic system angle to realize the track tracking of the unmanned vehicle.
In the above technical solution, the unmanned vehicle of repeated operation type includes, but is not limited to, an unmanned road roller for reciprocating rolling operation, a forklift for reciprocating cargo, a bulldozer or a loader for reciprocating operation.
In the above technical solution, the interference rejection controller for the distance error in step 2 includes a distance error feedback controller and a distance error disturbance suppressor, wherein in the distance error disturbance suppressor:
the extended state observer estimates the instant disturbance of the distance error in real time by using the actual course angle and the actual distance error information;
the iterative learner of the instant total disturbance of the distance errors performs iterative estimation on the accumulated disturbance of the distance errors in the control cycle according to the instant disturbance of the distance errors observed by the extended state observer and the accumulated disturbance of the distance errors obtained by the iterative learner of the instant total disturbance of the distance errors in the previous control cycle to obtain an accumulated disturbance estimation value of the distance errors;
model calculation for control a modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the distance error, the accumulated disturbance of the distance error and the modeling disturbance are compensated in real time in the distance error disturbance suppressor, and the target course angle of the required unmanned vehicle is calculated together with the distance error feedback controller, so that the track tracking distance error tends to zero.
In the above technical solution, the disturbance rejection controller of the course angle in step 3 includes a course angle feedback controller and a course angle disturbance suppressor, wherein in the course angle disturbance suppressor:
the extended state observer estimates the instant disturbance of the course angle in real time by using the actual steering wheel rotation angle and the actual course angle information;
the iterative learner of the instant total disturbance of the course angle performs iterative estimation on the accumulated disturbance of the course angle in the control cycle according to the instant disturbance of the observed course angle of the extended state observer and the accumulated disturbance of the course angle obtained by the iterative learner of the instant total disturbance of the course angle in the last control cycle to obtain an accumulated disturbance estimation value of the course angle;
model calculation for control the modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the course angle, the accumulated disturbance of the course angle and the disturbance which can be modeled are compensated in real time in the course angle disturbance suppressor, and together with the course angle feedback controller, the target course angle of the needed unmanned vehicle is calculated, so that the tracking error of the track tracking course angle tends to zero.
In the above technical solution, the distance error in the step 2 is resistedThe disturbance controller is
Wherein u outlp The result of (1) is assigned to the target course angle
The distance error feedback controller is
u 0,outlp Is a virtual control amount in the distance error control, b 0,outlp Is the gain of the control input
Is a measure of vehicle speed; wherein u is 0,outlp ＝k p,outlp (0-e d )，k p,outlp Is a proportional control coefficient, which is adjusted according to the speed requirement of the control process response, e d Is the actual distance error; the distance error disturbance suppressor in the step 2 is
Wherein
Obtaining an instant disturbance estimation value of the distance error through an extended state observer;
the accumulated disturbance estimation value of the distance error is obtained by an iterative learner of the instant total disturbance of the distance error;
a modelable disturbance is calculated for the model used for the control.
In the above technical solution, in the step 2
By expanding in the state observerCalculating an instant disturbance suppressor of the distance error, wherein the instant disturbance suppressor of the distance error is as follows:
In the above-mentioned technical solution, the air conditioner,
obtained by an iterative learner of instantaneous total perturbation of the distance error, which is:
where k is the number of the repeated working cycles, each complete working process being a cycle, α 1 As forgetting factor, gamma 1 Is a learning factor.
In the above technical solution, the anti-interference controller of the course angle in step 3 is
Wherein u is inlp The calculated result is assigned to the target steering wheel angle
The course angle feedback controller is
u 0,inlp Is a virtual control quantity in course angle control, b 0,inlp Is the gain of the control input; wherein
k p,inlp Is a proportional control coefficient, is adjusted according to the speed requirement of the response of the control process,
is the target course angle and is the target course angle,
is the actual course angle, b 0,inlp Gain input into the course control loop; the course angle disturbance suppressor in the step 3 is
Wherein
The real-time disturbance estimation value of the course angle is obtained by an extended state observer;
the accumulated disturbance estimation value of the course angle is obtained through an iterative learning device of the real-time total disturbance of the course angle;
a modelable disturbance is calculated for the model used for the control.
In the above technical solution, the
The method is obtained by calculating an instant disturbance suppressor of a course angle in an extended state observer, wherein the instant disturbance suppressor of the course angle is as follows:
In the above technical solution, the
The method is obtained by an iterative learner of the instantaneous total disturbance of the course angle, and the learner comprises the following steps:
where k is the number of the repeated working cycles, each complete working process being a cycle, α 2 Is a forgetting factor, gamma 2 Is a learning factor.
Compared with the prior art, the invention has the beneficial effects that:
1) By combining the disturbance suppression based on the model, the disturbance suppression based on the extended state observer and the disturbance suppression based on reinforcement learning, the adaptability of the algorithm to uncertainty such as vehicle state change, road condition change and the like can be remarkably improved.
2) The invention aims at a special scene of an unmanned vehicle with repeated operation, utilizes an iterative learning method to gradually induce deterministic and periodic interference from repeated uncertainty, and further compensates in a controller in real time. The control idea is expected to gradually improve the anti-interference capability of the controller and continuously improve the control effect.
3) The method has the advantages that the rapid suppression of very definite disturbance is realized by using the model-based disturbance rejection, the inaccuracy of the model compensation of the extended state observer is used, and the observation burden of the extended state observer is gradually relieved by using iterative learning. The three anti-interference modes are matched with each other, so that the dependence of an algorithm on a high-precision model is avoided, the problem that the iterative learning is slow in the suppression speed of sudden and random interference is solved, the problem that the observation speed of the extended state observer is limited under the low sampling frequency is compensated, and an organically matched whole is formed.
4) The invention uses the iterative learning algorithm to aim at the estimated disturbance of the extended state observer, but not the tracking error in the conventional iterative learning. The thought enables the iterative learning and the extended state observer to be in seamless butt joint, and the conflict between the two types of closed-loop controllers in the dynamic control process is avoided.
Drawings
Fig. 1 is a schematic view of the frame structure of the present invention.
Detailed Description
The present invention will be described in further detail with reference to specific examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Example 1
A repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning comprises the following steps:
step 1, calculating the actual distance error of the unmanned vehicle according to the target track of the unmanned vehicle.
Step 2, calculating a target course angle of the unmanned vehicle by the anti-interference controller of the distance error through the actual distance error obtained in the step 1, the instant disturbance of the distance error obtained by the extended state observer and the accumulated disturbance of the distance error obtained by the iterative learner of the instant total disturbance of the distance error, so that the track tracking distance error of the unmanned vehicle tends to zero;
and 3, calculating a target steering wheel rotating angle or a hydraulic system rotating angle of the unmanned vehicle by the anti-interference controller of the course angle through the target course angle obtained in the step 2, the instant disturbance of the course obtained by the extended state observer and the accumulated disturbance of the course angle obtained by the iterative learner of the instant total disturbance of the course angle, so that the actual course angle of the unmanned vehicle approaches the target course angle obtained in the step 2.
And 4, sending the target steering wheel angle or the hydraulic system angle obtained in the step 3 to a steering wheel angle or hydraulic system angle controller, and performing closed-loop control on the actual steering wheel angle or hydraulic system angle to realize the track tracking of the unmanned vehicle.
In this embodiment, the unmanned vehicle includes, but is not limited to, an unmanned road roller for reciprocating rolling operation, a forklift for reciprocating cargo, a bulldozer for reciprocating operation, or a loader for reciprocating operation.
Actual distance error in step 1 (e) d ) The calculation method comprises the following steps:
wherein (x) a ,y a ) Is the coordinate of the start point of the target track, (x) b ,y b ) The coordinates of the target track end point, and (x, y) the actual coordinates of the vehicle positioning point.
Example 2
The present embodiment further describes the disturbance rejection controller for the distance error in step 2.
The disturbance rejection controller for the distance error in the step 2 comprises a distance error feedback controller and a distance error disturbance suppressor, wherein in the distance error disturbance suppressor:
the extended state observer estimates the instant disturbance of the distance error in real time by using the actual course angle and the actual distance error information;
the iterative learner of the instant total disturbance of the distance errors performs iterative estimation on the accumulated disturbance of the distance errors in the control cycle according to the instant disturbance of the distance errors observed by the extended state observer and the accumulated disturbance of the distance errors obtained by the iterative learner of the instant total disturbance of the distance errors in the previous control cycle to obtain an accumulated disturbance estimation value of the distance errors;
model calculation for control a modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the distance error, the accumulated disturbance of the distance error and the modeling disturbance are compensated in real time in the distance error disturbance suppressor, and the target course angle of the required unmanned vehicle is calculated together with the distance error feedback controller, so that the track tracking distance error tends to zero.
Preferably, the interference rejection controller for the distance error in step 2 is
Wherein u is outlp Assigning the result of (A) to a target course angle
The distance error feedback controller is
u 0,outlp Is a virtual control quantity (unit is rad) in the distance error control, b 0,outlp Is the gain of the control input
Is a measure of vehicle speed (in m/s); wherein u 0,outlp ＝k p,outlp (0-e d )，k p,outlp Is a proportional control coefficient, and can be controlled according to the control processShould the speed and speed requirements be adjusted, it is recommended to use but not limited to bandwidth based parameter tuning method (Gao z. Scaling and bandwidth-based controller tuning. Proceedings of the American control reference.2006, 6 4989-4996) e d Is the actual distance error (in m); the distance error disturbance suppressor in the step 2 is
Wherein
The estimated value (unit is m/s) of the instant disturbance of the distance error is obtained by an extended state observer;
the accumulated disturbance estimation value (unit is m/s) of the distance error is obtained by an iterative learning device of the instantaneous total disturbance of the distance error;
the modelable disturbance (in m/s) is calculated for the model used for the control. As is well known to those skilled in the art, depending on the selection of a particular unmanned vehicle model. For passenger vehicles, kinematic or dynamic models can be selected, and the model can be recommended but not limited to books (Gong Jiang, ginger rock, xuwei. Unmanned vehicle model predictive control [ M ]]2014.), for the model of articulated vehicle, an article (C. A while to a used an insulated vehicle in an underseground timing is recommended but not limited to]// Proceedings 1999IEEE International Conference on Robotics and Automation (Cat. No.99CH36288℃) IEEE,1999,4: 3020-3025. Both models are well known to those skilled in the art, and the detailed process is not described in detail.
In said step 2
The method is obtained by calculating an instant disturbance suppressor of a distance error in an extended state observer, wherein the instant disturbance suppressor of the distance error is as follows:
The above-mentioned
Obtained by an iterative learner of instantaneous total perturbation of distance errors, the learner being:
where k is the number of the repeated working cycles, each complete working process being a cycle, α 1 The larger the value of the variable is, the higher the forgetting speed of the past data is, and the gamma is the forgetting factor 1 For the learning factor, the smaller the value of the variable, the slower the learning speed.
Different from the traditional closed-loop controller, the algorithm of the invention adopts an iterative learning algorithm, so that the calculation of the 'instant disturbance suppressor of the distance error in the extended state observer' of the last working cycle is needed to obtain
Accumulated disturbance estimation value of course angle obtained by iterative learner of 'sum' distance error instant total disturbance
"stored in a database.
Example 3
This example further illustrates the disturbance rejection controller of the heading angle in example 1.
The disturbance rejection controller of the course angle in the step 3 comprises a course angle feedback controller and a course angle disturbance suppressor, wherein in the course angle disturbance suppressor:
the extended state observer estimates the instant disturbance of the course angle in real time by using the actual steering wheel rotation angle and the actual course angle information;
the iterative learner of the instantaneous total disturbance of the course angle performs iterative estimation on the accumulated disturbance of the course angle in the control cycle according to the instantaneous disturbance of the course angle observed by the extended state observer and the accumulated disturbance of the course angle obtained by the iterative learner of the instantaneous total disturbance of the course angle in the last control cycle to obtain an accumulated disturbance estimation value of the course angle;
model calculation for control the modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the course angle, the accumulated disturbance of the course angle and the disturbance which can be modeled are compensated in real time in the course angle disturbance suppressor, and together with the course angle feedback controller, the target course angle of the needed unmanned vehicle is calculated, so that the tracking error of the track tracking course angle tends to zero.
The anti-interference controller of the course angle in the step 3 is
Wherein u is inlp The calculated result is assigned to the target steering wheel angle
The course angle feedback controller is
u 0,inlp Is a virtual control quantity in course angle control, b 0,inlp Is the gain of the control input; wherein
k p,inlp Is a proportional control coefficient, can be adjusted according to the speed requirement of the response of the control process,
for the target heading angle (in rad),
is the actual course angle (in rad), b 0,inlp The gain input in the course control loop needs to be adjusted, namely b is adjusted according to the requirement on the control effect 0,inlp (ii) a The course angle disturbance suppressor in the step 3 is
Wherein
The real-time disturbance estimation value (unit is rad/s) of the course angle is obtained by an extended state observer;
obtaining an accumulated disturbance estimation value of a course angle through an iterative learning device of the real-time total disturbance of the course angle (the unit is rad/s);
modelable perturbations (in rad/s) calculated for the model used for control. As is well known to those skilled in the art, depending on the selection of a particular unmanned vehicle model.
In said step 2
The method is obtained by calculating an instant disturbance suppressor of a course angle in an extended state observer, wherein the instant disturbance suppressor of the course angle is as follows:
wherein the content of the first and second substances,
is a measure of the heading angle and,
is X 1 Is estimated value
Derivative with respect to time, U 2 For course angle control loopControl input in, i.e. actual steering wheel angle
(in rad), can be fed back by a positioning measurement system,
and
all can pass through formula
Iteration is carried out to obtain;
β 1,inlp and beta 2,inlp In order to expand the gain of the state observer, a pole allocation method which is well known in the industry can be adopted for setting;
the estimation value (unit is rad/s) of the accumulated disturbance of the course angle is obtained from an iterative learner of the real-time total disturbance of the course angle;
the modelable disturbance (in rad/s) calculated for the model used for control is well known to those skilled in the art, depending on the selection of the particular unmanned vehicle model. For passenger vehicles, kinematic or kinetic models can be selected, recommended but not limited to books (Gong Wei, ginger rock, xuwei. Unmanned vehicle model predictive control [ M ]]2014.), for the model of articulated vehicle, an article (c.white to a used an annular vehicle in an underlying timing optimization) is recommended but not limited to]// Proceedings 1999IEEE International Conference on Robotics and Automation (Cat. No.99CH36288℃) IEEE,1999,4: 3020-3025. Both models are well known to those skilled in the art, and the detailed process is not described in detail.
Said
The method is obtained by an iterative learner of the instantaneous total disturbance of the course angle, and the learner comprises the following steps:
where k is the number of repeated work cycles, and each complete work cycle (e.g. one pass or one pass from the vehicle to the vehicle between the start and end) is a cycle, α 2 The larger the value of the variable is, the higher the forgetting speed of the past data is, and the gamma is the forgetting factor 2 For the learning factor, the smaller the value of the variable, the slower the learning speed. Note that: different from the traditional closed-loop controller, the algorithm of the invention adopts an iterative learning algorithm, so that the real-time disturbance suppressor of the course angle in the extended state observer of the last working cycle needs to be calculated
Accumulated disturbance estimation value of course angle obtained by iterative learner of 'sum' course angle instant total disturbance
"stored in a database.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (4)
1. The repeated operation type unmanned vehicle track tracking control algorithm based on total disturbance instant observation and iterative learning is characterized by comprising the following steps of:
step 1, calculating an actual distance error of the unmanned vehicle according to a target track of the unmanned vehicle;
step 2, calculating a target course angle of the unmanned vehicle by the anti-interference controller of the distance error through the actual distance error obtained in the step 1, the instant disturbance of the distance error obtained by the extended state observer and the accumulated disturbance of the distance error obtained by the iterative learner of the instant total disturbance of the distance error, so that the track tracking distance error of the unmanned vehicle tends to zero;
the interference rejection controller of the distance error in the step 2 is
Wherein u outlp Assigning the result of (A) to a target course angle
The distance error feedback controller is
u 0,outlp Is a virtual control amount in distance error control, b 0,outlp Is the gain of the control input
Is a measure of vehicle speed; wherein u is 0,outlp ＝k p,outlp (0-e d )，k p,outlp Is a proportional control coefficient, which is adjusted according to the speed requirement of the control process response, e d Is the actual distance error; the distance error disturbance suppressor in the step 2 is
Wherein
Obtaining an instant disturbance estimation value of the distance error through an extended state observer;
the accumulated disturbance estimation value of the distance error is obtained by an iterative learner of the instant total disturbance of the distance error;
a modelable disturbance calculated for the model used for the control; in said step 2
The method is obtained by calculating an instant disturbance suppressor of a distance error in an extended state observer, wherein the instant disturbance suppressor of the distance error is as follows:
wherein the content of the first and second substances,
Y 1 ＝e d，meas is a measure of the distance error and,
is X 1 Is estimated value
Derivative with respect to time, U 1 For control input in a range error control loop, i.e. actual heading angle
And
all can pass through formula
Iteration is carried out to obtain;
β 1,outlp and beta 2,outlp To expand the state observer gain;
obtaining an accumulated disturbance estimation value of the distance error from an iterative learner of the instant total disturbance of the distance error;
modelable disturbances calculated for the model used for the control;
step 3, the disturbance rejection controller of the heading angle calculates a target steering wheel corner or a hydraulic system corner of the unmanned vehicle through the target heading angle obtained in the step 2, the instant disturbance of the heading obtained by the extended state observer and the accumulated disturbance of the heading angle obtained by the iterative learner of the instant total disturbance of the heading angle, so that the actual heading angle of the unmanned vehicle approaches to the target heading angle obtained in the step 2;
the anti-interference controller of the course angle in the step 3 is
Wherein u is inlp Is assigned to the target steering wheel angle
The course angle feedback controller is
u 0,inlp Is a virtual control quantity in course angle control, b 0,inlp Is the gain of the control input; wherein
k p,inlp Is a proportional control coefficient, is adjusted according to the speed requirement of the response of the control process,
is the target course angle and is the target course angle,
is the actual course angle, b 0,inlp Gain input in the course control loop; the course angle disturbance suppressor in the step 3 is
Wherein
The instantaneous disturbance estimated value of the course angle is obtained through an extended state observer;
the accumulated disturbance estimation value of the course angle is obtained by an iterative learner of the real-time total disturbance of the course angle;
a modelable disturbance calculated for the model used for the control;
said
Calculated by an instantaneous disturbance suppressor of the course angle in the extended state observer, said flightThe instant disturbance suppressor for the steering angle is:
wherein, the first and the second end of the pipe are connected with each other,
is a measure of the heading angle and,
is X 1 Is estimated value
Derivative with respect to time, U 2 For control input in a course angle control loop, i.e. actual steering wheel angle
Can be obtained by the feedback of the positioning measurement system,
and
all can pass through formula
Iteration is carried out to obtain;
β 1,inlp and beta 2,inlp To expand the state observer gain;
iterative learner of instantaneous total perturbation from course angle for accumulated perturbation estimate of course angleObtaining;
a modelable disturbance calculated for the model used for the control;
said
The method is obtained through an iterative learner of the instantaneous total disturbance of the course angle, and the learner is as follows:
where k is the number of the repeated working cycle, each complete working process is a cycle, α 2 Is a forgetting factor, gamma 2 Is a learning factor;
and 4, sending the target steering wheel angle or the hydraulic system angle obtained in the step 3 to a steering wheel angle or hydraulic system angle controller, and performing closed-loop control on the actual steering wheel angle or hydraulic system angle to realize the track tracking of the unmanned vehicle.
2. The repetitive-work unmanned vehicle trajectory tracking control algorithm based on total disturbance instantaneous observation and iterative learning of claim 1, wherein the repetitive-work unmanned vehicle comprises but is not limited to an unmanned road roller for reciprocating milling work, a forklift for reciprocating cargo, a bulldozer for reciprocating work, or a loader.
3. The repetitive-task unmanned vehicle trajectory tracking control algorithm based on total disturbance instantaneous observation and iterative learning of claim 1, wherein the immunity controller for distance error in step 2 comprises a distance error feedback controller and a distance error disturbance suppressor, wherein in the distance error disturbance suppressor:
the extended state observer estimates the instant disturbance of the distance error in real time by using the actual course angle and the actual distance error information;
the iterative learner of the instant total disturbance of the distance errors performs iterative estimation on the accumulated disturbance of the distance errors in the control cycle according to the instant disturbance of the distance errors observed by the extended state observer and the accumulated disturbance of the distance errors obtained by the iterative learner of the instant total disturbance of the distance errors in the previous control cycle to obtain an accumulated disturbance estimation value of the distance errors;
model calculation for control the modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the distance error, the accumulated disturbance of the distance error and the modeling disturbance are compensated in real time in the distance error disturbance suppressor, and the target course angle of the required unmanned vehicle is calculated together with the distance error feedback controller, so that the track tracking distance error tends to zero.
4. The repetitive-operation unmanned vehicle trajectory tracking control algorithm based on total disturbance instantaneous observation and iterative learning of claim 1, wherein the disturbance rejection controller for the course angle in step 3 comprises a course angle feedback controller and a course angle disturbance suppressor, wherein in the course angle disturbance suppressor:
the extended state observer estimates the instant disturbance of the course angle in real time by using the actual steering wheel rotation angle and the actual course angle information;
the iterative learner of the instantaneous total disturbance of the course angle performs iterative estimation on the accumulated disturbance of the course angle in the control cycle according to the instantaneous disturbance of the course angle observed by the extended state observer and the accumulated disturbance of the course angle obtained by the iterative learner of the instantaneous total disturbance of the course angle in the last control cycle to obtain an accumulated disturbance estimation value of the course angle;
model calculation for control a modelable disturbance forms a model-based disturbance feedforward;
therefore, the instant disturbance of the course angle, the accumulated disturbance of the course angle and the disturbance which can be modeled are compensated in real time in the course angle disturbance suppressor, and together with the course angle feedback controller, the target course angle of the needed unmanned vehicle is calculated, so that the tracking error of the track tracking course angle tends to zero.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010140701.0A CN113341943B (en) | 2020-03-03 | 2020-03-03 | Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010140701.0A CN113341943B (en) | 2020-03-03 | 2020-03-03 | Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113341943A CN113341943A (en) | 2021-09-03 |
| CN113341943B true CN113341943B (en) | 2022-10-11 |
Family
ID=77467402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010140701.0A Active CN113341943B (en) | 2020-03-03 | 2020-03-03 | Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113341943B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106208824A (en) * | 2016-07-22 | 2016-12-07 | 浙江工业大学 | Multi-motor synchronous control method based on active disturbance rejection iterative learning |
| CN110209177A (en) * | 2019-07-02 | 2019-09-06 | 天津大学 | Pilotless automobile control system and method based on model prediction and active disturbance rejection |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6786407B2 (en) * | 2017-01-23 | 2020-11-18 | 株式会社クボタ | Work vehicle wireless management system |
| CN107991867A (en) * | 2017-11-28 | 2018-05-04 | 浙江工业大学 | Iterative learning contour error control method of networked multi-axis motion control system based on active disturbance rejection controller |
| CN108303877B (en) * | 2018-01-10 | 2020-05-05 | 浙江工业大学 | ESO-based networked motion control system iterative learning high-precision contour tracking control method |
| CN108873906A (en) * | 2018-07-10 | 2018-11-23 | 北京踏歌艾尔机器人科技有限公司 | Roll the unmanned control method of engineering truck and device and readable storage medium storing program for executing |
| CN109542105A (en) * | 2018-12-28 | 2019-03-29 | 北京龙马智行科技有限公司 | A kind of subgrade and pavement Intelligent unattended rcc system |
| CN110531777B (en) * | 2019-08-13 | 2020-11-10 | 北京理工大学 | Four-rotor aircraft attitude control method and system based on active disturbance rejection control technology |
| CN110764499A (en) * | 2019-09-27 | 2020-02-07 | 南京航空航天大学 | Device and method for ensuring straight-line walking of mower |
-
2020
- 2020-03-03 CN CN202010140701.0A patent/CN113341943B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106208824A (en) * | 2016-07-22 | 2016-12-07 | 浙江工业大学 | Multi-motor synchronous control method based on active disturbance rejection iterative learning |
| CN110209177A (en) * | 2019-07-02 | 2019-09-06 | 天津大学 | Pilotless automobile control system and method based on model prediction and active disturbance rejection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113341943A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fang et al. | Trajectory tracking control of farm vehicles in presence of sliding | |
| Peng et al. | Leader–follower formation control of nonholonomic mobile robots based on a bioinspired neurodynamic based approach | |
| Samuel et al. | A review of some pure-pursuit based path tracking techniques for control of autonomous vehicle | |
| De Simone et al. | Identification and control of a unmanned ground vehicle by using Arduino | |
| CN110780594B (en) | Path tracking method and system of intelligent vehicle | |
| Arogeti et al. | Path following of autonomous vehicles in the presence of sliding effects | |
| CN109857100B (en) | Composite track tracking control algorithm based on inversion method and fast terminal sliding mode | |
| Keymasi Khalaji | Modeling and control of uncertain multibody wheeled robots | |
| Jiao et al. | A novel aircraft anti-skid brake control method based on runway maximum friction tracking algorithm | |
| Ribeiro et al. | Nonlinear model predictive visual path following control to autonomous mobile robots | |
| Low et al. | Maneuverability and path following control of wheeled mobile robot in the presence of wheel skidding and slipping | |
| Hung et al. | Design of a sliding mode controller for an automatic guided vehicle and its implementation | |
| Velasco-Villa et al. | Trajectory tracking for awheeled mobile robot using a vision based positioning system and an attitude observer | |
| Ding et al. | Composite SOSM controller for path tracking control of agricultural tractors subject to wheel slip | |
| Zhao et al. | Nonlinear model predictive control-based guidance algorithm for quadrotor trajectory tracking with obstacle avoidance | |
| Rigatos et al. | A nonlinear H-infinity control approach for autonomous truck and trailer systems | |
| Rigatos | A nonlinear optimal control approach for tracked mobile robots | |
| Spitzer et al. | Feedback linearization for quadrotors with a learned acceleration error model | |
| Yuan et al. | Differential flatness-based adaptive robust tracking control for wheeled mobile robots with slippage disturbances | |
| CN113341943B (en) | Repeated operation type unmanned vehicle trajectory tracking control algorithm based on total disturbance instant observation and iterative learning | |
| Zou et al. | An adaptive control strategy for indoor leader-following of wheeled mobile robot | |
| Ding et al. | Straight-line tracking controller design of agricultural tractors based on third-order sliding mode | |
| Kim et al. | Study on vehicle lateral control for backward driving | |
| CN114063621B (en) | Wheel type robot formation tracking and obstacle avoidance control method | |
| CN114690754B (en) | Repeated operation type unmanned vehicle iterative learning self-optimizing tracking control algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |