Mine-used bolter is the main equipment to solve the imbalance of excavation and anchor in well mining, and the manipulator is the main working mechanism of mine bolt drilling rig. The manipulator positioning requires high rapidity and stability. For this reason, this paper proposes a composite control method of “input shaping + fractional order PDμ control”. According to the mathematical model of the valve-controlled cylinder, the fractional-order controller PDμ is developed. At the same time, the input shaping is used to feed forward the accurate positioning and path planning of the manipulator, which not only improves the robustness of the system, but also shortens the stability time of the system and restrains the maximum amplitude of the system vibration. In this paper, the control effects of fractional order PDμ controller and integer order PD controller are compared. The results show that the maximum amplitude of the control system is reduced by 75% and the stabilization time is reduced by 60% after using the fractional order PDμ controller, which fully reflects the superiority of the fractional order controller in response speed, adjusting time, and steady-state accuracy. Finally, the control effects of “input shaping + fractional order PDμ control” and fractional order PDμ controller on the stability of the system are compared. The maximum amplitude of the system was reduced by 50% by using “input shaping + fractional order PDμ control”. Numerical simulation confirms the feasibility and effectiveness of the composite control method. This composite control method provides theoretical support for the precise positioning of the manipulator, and the high stability and high safety of the manipulator also expand the application scope and depth of the composite control method.
1. Introduction
At present, underground mining areas mainly require excavation and support, but there is a serious imbalance between them. The artificial support time accounts for 2/3 of the mining time. In order to improve the supporting efficiency, a mine-used bolter is developed. A mine-used bolter consists of a chassis, a variable frequency electric traction fixed interval control system, an electric control system, two manipulators which have the function of leveling telescopic, two automatic drilling frames, a driving operation mechanism, fore and aft stabilizers, a roll-cable device, a spread roof mesh device, an auto spraying device, etc. The schematic diagrams of the mine-used bolter are shown in Figures 1 and 2. The manipulator is the main working mechanism of the device, consisting of 6 free joints. The position control performance of each joint is directly related to the support efficiency of the mine-used bolter, so controlling a mine-used bolter’s manipulator has become the hotspot and difficulty of research.
Schematic diagram I of mine-used bolter structure. 1: frequency control of motor speed walking mechanism, 2: electrical control system, 3: telescopic arm, 4: automatic drilling frame, 5: hydraulic system, 6: frame, 7: driving operation and monitor, and 8: directional cruise control and step measurement system.
Schematic diagram II of mine-used bolter structure. 9: fore and aft stabilizer, 10: leveling mechanism, 11: spread roof mesh device, and 12: spraying volume box.
Previously, a traditional PID control was used for the spatial position and attitude of the manipulator. In view of the obvious advantages of the PID controller [1–3], the preview following theory [4, 5] has achieved good results in traditional PID control. However, when the mine-used bolter works underground, the underground environment is harsh and the external load is prone to change, and the load will change with the different needs of the underground operators. Especially under special conditions, the control effect of traditional PID is unsatisfactory if the manipulator needs to achieve the goal of track tracking and spatial positioning quickly. At this time, the parameters of the PID need to be reset. Even after adjusting the parameters, the expected target value of the control effect of the mine-used bolter could not be reached by the integer order PID controller. In order to improve the control effect of integer order PID controller, the authors Ge LI, Wang Y F, and Liu Land so on in [6–16] used different PID algorithms in various control systems, such as integral separation PID control algorithm, fuzzy control PID control algorithm, neural network PID control algorithm, and so on. In essence, none of the above algorithms overcome the shortcomings of the traditional PID algorithm with relatively limited parameter range. According to the above discussion, a fractional order PIλDμ controller with better dynamic and robustness must be developed to provide a greater guarantee for the spatial positioning of the manipulator of the mine anchor drill underground. Podlubny [17] first proposed the model of fractional order proportional-integral-differential controller. The basic knowledge of the definition, properties, Laplace transformation, and application of fractional order calculus is introduced in [18–22]. Compared with the integer PID, the expression of fractional PIλDμ controller has two parameters λ,μ, which increases the adjustment range of parameters. It provides the theoretical basis and technical support for the formulation of fractional order proportional-integral-differential controller for the position and attitude of the manipulator of the mine-used bolter. References [23, 24] also put forward optimization methods for two parameters in fractional order PIλDμ expression. The common optimization methods are combined with neural network, particle swarm optimization, and genetic algorithm based on finite difference to find the optimal parameters of λ,μ, which ensures the optimal control effect of fractional order PIλDμ controller. However, all the above research work was based on the design and optimization of a single fractional order PIλDμ controller and did not combine any kinds of feed-forward control-input shaping devices to form a composite controller to control the stability of the system together. In [25], a composite control method based on proportional-differential feedback and moment feed-forward was proposed to achieve the joint attitude of the mechanism. Similarly, a hybrid control method, which combined variable structure control and input shaping technology, was proposed in document [26] and applied to fast and agile maneuvering control of satellite attitude. The above two references [25, 26] showed that compound control could not only effectively improve the anti-jamming performance of the system, but also improve the speed of trajectory tracking. In view of the advantages of fractional order algorithm and composite control method, a composite control method combining fractional order PDμ controller and input shaping technology is developed for the space trajectory tracking and precise positioning of the manipulator of mine-used bolter, which ensures that the manipulator can efficiently, safely, and steadily complete the trajectory tracking and positioning while working underground. The control block diagram is shown in Figure 3.
Schematic diagram of control system structure.
2. Mathematical Model of Electrohydraulic Proportional Control System for Manipulator
The rotation angle and displacement of each joint of the manipulator of the mine-used bolter are driven by the corresponding proportional valve and cylinder. In order to realize the trajectory tracking and precise positioning of the manipulator of the mine-used bolter, the electrohydraulic proportional position control system is adopted at each joint. Figure 4 is a block diagram of the transfer function of a position system for the electrohydraulic proportional control of a hydraulic cylinder with a valve controlled single pole. The transfer function of the hydraulic power system can be obtained from Figure 4.
Block diagram of transfer function of location system.
The external load is F=0, Rs which is a reference input, and Xp is the transfer function of the output of the hydraulic cylinder. The process parameters are shown in Table 1. The relevant parameter values are subdivided into the above sections. In the case of neglecting the external load force, the open-loop transfer function of each joint position system of the manipulator can be solved, as shown in (1) formula.(1)Gs=kq/Apss2/ωh2+2ξh/ωhs+1=0.155/3434×10-6ss2/20.92+2×0.4/20.9s+1=45.1ss2/436.81+0.04s+1,where ωh=4βeAp2/Vtmt (ωh is natural frequency of hydraulic cylinder, unit: rad/s), ξh=0.4 (ξh is hydraulic damping ratio), Ap is effective working area of piston hydraulic cylinder (unit: m2), and kq is proportional valve flow gain.
Parameters of valve controlled hydraulic cylinder.
Parameters of the symbol
Parameter name
Unit
Oil cylinder
L
trip
mm
1100
D
cylinder bore
mm
80
d
piston rod diameter
mm
45
ξh
hydraulic damping ratio
--
0.4
mt
drive equivalent load
kg
2000
βe
oil equivalent elastic modulus
Pa
7.0∗108
cϖ
proportional valve core area gain
mm
18
Δp
proportional valve rated core inlet and outlet differential pressure
Mp
1.6
ρ
the oil density
kg/m3
860
3. Design of Fractional Order PDμ Controller for the Manipulator of Mine-Used Bolter
The transfer function of fractional order PIλDμ is as follows:(2)Gs=Kp1+Kisλ+Kdsμ
When the parameters λ and μ are taken different values, the PIλDμ controllers of different fractional stages can be obtained. The range of parameters is wider and the system regulation is more flexible.
The main design methods of PIλDμ controller are dominant pole method [25, 26], optimization methods [27], and so on. In order to track and locate the space trajectory of the manipulator accurately, the fractional order PIλDμ controller is designed by using the method of amplitude, phase margin, and robustness constraints.
3.1. Design of a Fractional Order PDμ Controller
Eq. (2) indicates that the transfer function of the fractional order PDμ controller is(3)Cs=Kp1+Kdsμ,
Kp refers to a proportional constant, Kd refers to a differential constant, and u refers to the fractional order of the controller, μ∈0,1. The phase-frequency and amplitude characteristics of the valve-controlled oil cylinder for the manipulator are as follows:(4)argGjω=-π2-arctan2ξh/ωhω1-ω2/ωh2,(5)Gjω=kq/Aω1-ω2/ωh22+2ξh/ωhω.Eq. (3) of the PDμ fractional order controller can be shown as Laplace transform as follows:(6)Cjω=KpKdjωμ=Kp1+Kdωμcosμπ2+jKdωμsinμπ2.
The phase-frequency and amplitude characteristics are(7)argCjω=arctansin1-μπ/2+Kdωμcos1-μπ/2-1-μπ2,(8)Cjω=Kp1-Kdωμcosμπ22+Kdωμsinμπ22.
The open-loop transfer function is(9)Ls=CsGs.
The phase characteristics of the open-loop transfer function can be obtained through (4) and (7) as follows:(10)argLjω=arctansin1-μπ/2+Kdωμcos1-μπ/2-1-μπ2-π2-arctan2ξh/ωhω1-ω2/ωh2
The fractional order PDμ controller is analyzed in frequency domain, which must satisfy three constraints. It can be derived from the basic definition of cut-off frequency, amplitude, and phase margin.
(1) Constraint of the Phase Margin(11)argLjωcg=argCjωcgGjωcg=-π+φm.
ϕm refers to the phase margin and ωcg refers to the cut-off frequency. The phase margin can be regarded as a phase change that can be increased before the system enters a stable state. If the phase margin is large, then the system is stable. However, the time response speed decreases at the same time. Therefore, an appropriate phase margin must be selected based on different controlled objects.
(2) Robustness Constraint(12)dargLjωdωω=ωcg=0.
The derivative is obtained for the phase–frequency function of the open-loop system Ljω based on (12), and its zero point is fixed at the cut-off frequency. Thus, the phase becomes flat around the cut-off frequency ωcg, and the closed-loop system is robust to changes in system gain. Even when the system gain is changed in a certain range, the system overshoot remains unchanged. This constraint condition is the core reason that the manipulator has good robustness.
(3) Constraint of the Amplitude Margin(13)Ljωcg=CjωcgGjωcg=1.
The parameters of the fractional order PDμ controller can be obtained by the constraint of the amplitude margin.(14)argLjω=arctansin1-μπ/2+Kdωμcos1-μπ/2-1-μπ2-π2-arctan2ξh/ωhω1-ω2/ωh2=-π+ϕm.
The relationship between Kd and μ can be obtained using (14) as follows:(15)Kd=1ωcgμtanϕm+arctanωcg2ξh/ωh1-ω2/ωh2-μπ2cos1-μπ2-1ωcgμsin1-μπ2.
It can be obtained by the robustness constraint.(16)dargLjωdωω=ωcg=Kdμωμ-11+sin1-μπ+Kdωμ/cos1-μπ/22+-2ξh1+2ξh/ωhω/1-ω2/ωh21/ω2+1/ωh21/ω-ω/ωh22=0.
The Kd function can be obtained using (16).(17)AKd2+BKd+C=0,(18)Kd=-B±B2-4AC2A,(19)A=2ξhωh2+ω2ωh2ωcg2μ,B=4ξhωh2+ω2ωh2sin1-μ2ωcgμ-μωμ+1ωh2cos1-μ2C=2ξhωh2+ω2ωh2cos21-μ2+2ξhωh2+ω2ωh2sin21-μ2
It can be obtained by using the constraint of the amplitude margin.(20)Ljωcg=CjωGjω=Kp1+Kdωμcosμλ22+Kdωμsinμλ22Kq/Aω1-ω2/ωh2+2ξh/ωhω=1.
Three parameters required by the expression of the PDμ controller, namely, Kp=32,Kd=0.189,μ=0.308 are obtained with the aid of Matlab-Simulink. The expression of the fractional order PDμ controller can then be obtained.(21)Cs=321+0.189s0.308.
4. Input Shaping Technology
Input shaping [28, 29] refers to the convolution of a pulse series with a periodic expected input, and the resulting instruction is used as the actual control instruction of the system to drive the system. Among them, ZD, ZVD, ZVDD, EI, and EI-two hump shaper are widely used in forming [30]. The calculation parameters of each shaper can be calculated by document [31]. The method is applicable to rigid and elastic vibration of the system. The parameters of the input shaper can be suppressed. The input shaper is a pure lag feedforward unit composed of a different gain and the same time interval. Its mathematical expression is(22)Is=∑i=1mAie-i-1sT,m=2,3⋯.
Ai is the gain, T is the time interval, and m is the number of gains contained in the molding device. The above equation is described as time domain:(23)Is=∑i=1mAie-i-1sT,m=2,3⋯.
Equation (23) represents the pulse sequence with the amplitude Ai of the first pulse as the time zero and the time interval as T. The input instruction is denoted as Rt, and the molding instruction is denoted as Ut; thus it can be seen that the convolution relation of formula (24) is valid.(24)Ut=Rt∗It.
Rt can be any kind of incentive signal.
In order to realize the precise positioning and trajectory tracking of the manipulator in the underground space, this paper adopts the three-pulse input shaping tool ZVD as the forward feedback controller. The second-order closed-loop transfer function of the manipulator is as follows:(25)Ps=CsGs1+CsGs=kq/Ap·kp1+kds/ss2/ωh2+2ξh/ωhs+11+kq/Apkp1+kds/ss2/ωh2+2ξh/ωhs+1=kq/Ap·kp1+kdsss2/ωh2+2ξh/ωhs+1+kq/Ap·kp1+kds=1444×1+0.189sss2/436.81+0.04s+1+1444×1+0.189s=1.038s2/119585+12.1978/119585s+1.Structure of three-pulse ZVD input shaping is(26)C2=∑i3Aie-tis=A1e-t1s+A2e-t2s+A3e-t3s
5. Modeling and Simulation of the Control System5.1. Modeling of the PDμ Control System
Under the environment of Matlab-Simulink, the fractional order PDμ controller is used to control the valve-controlled cylinder of the space position and attitude actuator of the manipulator of the mine-used bolter. The unit step signal is used as input signal, and the valve-controlled cylinder at each joint position of the manipulator of the mine-used bolter is used as a control object. The control model is set up as shown in Figure 5.
Manipulator of mine-used bolter “input shaping + fractional order PDμ controller” system model.
5.2. Simulation of the Control System
The control effects of fractional order PDμ controller and integer order PD controller on system stability are compared, as shown in Figure 6. Finally, a compound control method, i.e., input shaper + fractional order PDμ control, is adopted to apply on the control system. The stability of the system is numerically analyzed and compared with that of a single fractional order PDμ controller, as shown in Figure 7.
Comparison of unit step response between fractional order PDμ controller and integer order PD controller.
Comparison of unit step response between fractional order PDμ controller and “input shaping + fractional order PDμ”.
Figure 6 compares the unit step response of a fractional PDμ controller and that of an integer order PD controller. Figure 6 shows that a fractional order controller PDμ has a better control effect than an integer order PD controller in response speed, adjustment time, and steady-state accuracy. The simulation results show that, under the influence of a fractional order PDμ controller, the maximum amplitude of the control system is 1.4, and the stability time is 0.055s; under the influence of an integer order PD controller, the maximum amplitude is 1.1, and the stability time is 0.022s. Through numerical simulation comparison, it can be seen that the maximum amplitude of the control system is reduced by 75% compared with the stability value 1 of the system, and the stability time is reduced by 60%, which fully reflects the superiority of the fractional order PDμ controller.
The step response of fractional order PDμ controller is compared with that of “input shaping + fractional order PDμ control”, as shown in Figure 7. It can be seen that the control system quickly tends to be stable when the compound control “input shaping device + fractional order PDμ controller" is adopted. In particular, the maximum amplitude decreases significantly from 1.5 to 1.25. Compared with the stability value 1, the maximum amplitude decreases by 50%, and the control system quickly tends to the stable stage. Underground, the reduction of maximum amplitude is of great significance to the work of the manipulator of a mine-used bolter and provides guaranteed accuracy and safety of roof or side support. The high precision positioning of the manipulator could avoid the interference between the equipment and the coal wall and improves the safety of underground workers and equipment.
In the environment of Matlab-Simulink, the simulation model of the mathematical models of controlled objects and the composite controllers are built in this section, which proves the effectiveness and superiority of the method.
6. Conclusions
Taking the manipulator of mine-used bolter as the research object, in order to realize the space trajectory tracking and accurate positioning of the manipulator, the following work has been done.
(1) The mathematical model of the electrohydraulic proportional control system of the manipulator was solved, and the PDμ controller was formulated based on a fractional order algorithm. Based on the definition and classification of input shaping technology, a ZVD feedforward controller was designed.
(2) Under the Matlab-Simulink environment, the control system model was built and numerical simulation was carried out. The fractional order PDμ controller and the integer order PD controller were compared with the control effect of the system. Under the influence of a fractional order PDμ controller, the maximum amplitude of the control system was reduced by 75%, and the stabilization time was reduced by 60%. It is verified that a fractional order PDμ controller has superiority over an integer order PD controller in response speed, adjusting time and steady-state accuracy.
(3) Finally, the control effects of “input shaping + fractional order PDμ controller” and “input shaping + fractional order PDμ controller” on system stability were compared. With the “input shaping + fractional order PDμ control”, the maximum amplitude of the system was significantly reduced by 50%. In underground operation, the reduction of the maximum amplitude can effectively improve the positioning accuracy of the end of the manipulator, ensure the supporting effect of the roadway, and further improve the safety of equipment and personnel. The method of “input shaping + fractional order PDμ control” developed in this paper can restrain the oscillation of the manipulator of the mine-used bolter in the process of trajectory tracking and precise positioning, so that the control system has strong anti-interference ability and good robustness. The composite control method provides a theoretical support for the precise positioning of the manipulator. At the same time, the high stability and high safety of the manipulator of the mine-used bolter also expand the scope and depth of the application of the composite control method.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This work was supported by Project Name: Research on automatic loading technology of mechanical anchorage agent; Project Number: 2018-TD-MS049; Project Name: Two arm anchor cable anchor car; Project Number: 2018-TD-MS047; Project Name: Influence of cooling strength distribution of casting roll on quality of side of magnesium alloy strip; Project Category: Education innovation project for graduate students of key technical materials and basic parts of clean energy and modern transportation equipment; Project Name: Development of self-adaptive synchronous tail without secondary support in driving face of medium broken roof; Project Number: 2018-TD-MS044.
AngK. H.ChongG.LiY.PID control system analysis, design, and technology20051345595762-s2.0-2244444762310.1109/TCST.2005.847331ÅströmK. J.HägglundT.The future of PID control20019111163117510.4173/mic.1994.3.1MR1307091RiveraE. D.MorariM.SkogestadS.Internal model control: PID controller design198625121632163LimingW.2007Beijing, ChinaScience PressDellerJ. R.HansenJ. H. L.ProakisJ. G.1993Macmillan Publishing Company10.1109/9780470544402JiuhongY.XiaozengW.Resistance furnace temperature control system based on integral separation PID algorithm20122016669GeL. I.JiaY.ZhangH.Application of integral-separation PID control algorithm in PLC-based tension control system2008298109112LinR.YangF.QiuG.The neuron controller based on integral separation PID control algorithm2004ChenH. Z.Research of compress anti-surge based on the integral separation PID algorithm201118919356757010.4028/www.scientific.net/AMR.189-193.567ZhimeiZ.LiyingC.YihengZ.Blind-guiding robot research based on fuzzy PID control algorithm20153318185SahuB. K.PatiS.MohantyP. K.PandaS.Teaching-learning based optimization algorithm based fuzzy-PID controller for automatic generation control of multi-area power system2015272402492-s2.0-8491582065310.1016/j.asoc.2014.11.027WangY. F.Design and research of time optimal fuzzy-PID control algorithm200484366365LiuL.LuoJ.Research of PID control algorithm based on neural network2011136988699310.1016/j.egypro.2011.12.409JahediG.ArdehaliM. M.Genetic algorithm-based fuzzy-PID control methodologies for enhancement of energy efficiency of a dynamic energy system201152172573210.1016/j.enconman.2010.07.051ZhaoJ.HanL.WangL.YuZ.The fuzzy PID control optimized by genetic algorithm for trajectory tracking of robot arm201655655910.1109/WCICA.2016.7578443Nian-XiangW. U.Research on optimization of tracking control strategy of moving robot based on fuzzy PID control algorithm2016PodlubnyI.Numerical solution of ordinary fractional differential equations by the fractional difference methodpp. 507-515, 1997MR1638518XieJ.HuangQ.YangX.Numerical solution of the one-dimensional fractional convection diffusion equations based on Chebyshev operational matrix201651, article no. 114910.1186/s40064-016-2832-y2-s2.0-84979776383XieJ.HuangQ.ZhaoF.GuiH.Block pulse functions for solving fractional Poisson type equations with Dirichlet and Neumann boundary conditions201713210.1186/s13661-017-0766-0MR3622825XieJ.YaoZ.WuR.DingX.ZhangJ.Block-pulse functions method for solving three-dimensional fractional Poisson type equations with Neumann boundary conditions201812610.1186/s13661-018-0945-7MR3769839XieJ.YaoZ.GuiH.A two-dimensional Chebyshev wavelets approach for solving the Fokker-Planck equations of time and space fractional derivatives type with variable coefficients201833219720810.1016/j.amc.2018.03.040ZhangG. W.Hong-YanA. M.QiuR.Application of fuzzy PID control algorithm in curtain system2014DabiriA.MoghaddamB. P.MachadoJ. A. T.Optimal variable-order fractional PID controllers for dynamical systems201810.1016/j.cam.2018.02.029MR3787676LiangT.ChenJ.LeiC.Algorithm of robust stability region for interval plant with time delay using fractional order PIλDμ controller2012172979991WeiminZ.LiaoyingY. Y.An agile satellite rapid attitude maneuver control method using input modeling20162542732YangX.CuiJ.LaoD.LiD.ChenJ.Input shaping enhanced active disturbance rejection control for a twin rotor multi-input multi-output system (TRMS)20166228729810.1016/j.isatra.2016.02.0012-s2.0-84959088832PetrasI.The fractional-order controllers: Methods for their synthesis and application1999509-10284288JinpengY.PolyesterY.2006BBSRomanoM.AgrawalN. B.BernellizazzeraF.Experiments on command shaping control of a manipulator with flexible links200225223223910.2514/2.4903ZhongchaoW.Research on attitude maneuver method of large flexible satellite input forming based on PD control20173428598ZhangT.LinK.ZouY.An input shaper with control of error optimization for end-effector's residual vibration suppression of industrial robots2018