Sage Journals: Discover world-class research

Abstract

At present, marine resources, especially the deep-sea resources, are becoming more and more important in resource exploitation globally, and hence, widely used deep-sea cranes are playing essential roles. For such systems, the bridge frames and trolleys are set up above the water, while payloads are transported under the water. In this underwater situation, there exist hydrodynamic forces such as complicated disturbances to the crane systems, making the payload vibration and rope flexibility more obvious. For the sake of improving working efficiency, considering the constraints of all the state variables, an anti-vibration trajectory is designed for the trolley motion, which can not only ensure trolley positioning but also suppress the flexible payload’s vibrations. Then, the state variables are constrained within preset safety ranges. Finally, numerical simulation results prove the satisfactory performance of the designed method.

Keywords

Underactuated systems distributed-parameter systems flexible payload trajectory planning vibration control

Introduction

In these years, as powerful equipment, cranes have been widely applied in architectonics, nuclear power construction, petrochemical industry, port construction, industrial environmental protection, and so on. Therefore, crane systems^1–12 are playing significant roles in modern industry gradually.

According to various practical applications, cranes can be divided into rotary cranes, tower cranes, gantry cranes, overhead cranes, and so on. In the light of the working environment, there are cranes operating on land and under water. Although their specific models and characteristics are different, they all belong to typical underactuated systems with more degrees of freedom (DOFs) than their independent control inputs.^13–22 Underactuated systems have stronger nonlinearities with variables being coupled with each other, making it more difficult to analyze and solve their control problems compared with fully actuated systems. However, in terms of energy saving, structural complexity, and cost reduction, underactuated systems are superior to fully actuated ones. Aside from cranes, translational oscillations with a rotational actuator (TORA)^23–25 systems and underactuated robot^26–28 systems also belong to underactuated systems.

An underwater crane system, as the term suggests, refers to a type of crane whose driving part is set above water with its flexible payload being transported in water. At present, the crane system is greatly used in oil exploration, sea resources exploitation, nuclear reactor, and other fields. In view of the flexible payload and high-precision requirements, the crane’s control objectives are to rapidly and precisely transport the payload to the preset position and also effectively restrict the vibration of the payload throughout the entire transportation process, bringing about zero residual vibration at the end of the process. There are two obvious problems that need to be addressed during the payload transportation. On one hand, the vibration of the payload under water will be influenced by the amplification of the hydrodynamic force compared with the one on land, which may cause large-amplitude vibration. On the other hand, in practice, cranes need to be manually operated by operators who are experienced and skilled, but minor operational mistakes caused by manual operation are inevitable, thus increasing the possibility of accidents. Consequently, considering the above two problems, it is demanded to devise a control approach to realize the goal of instantly reaching the desired position and effectively suppressing the residual vibration of the flexible payload simultaneously.

The underactuated systems’ control problems have been studied in academia, and many researchers have made remarkable achievements in recent decades. At present, the two most common existing methods are open-loop control strategies and closed-loop control methods. Specifically, open-loop control strategies include trajectory planning,²⁹ input shaping,^30,31 and feedforward control.³² For closed-loop control, there are also many representative approaches such as sliding mode control,^33–35 fuzzy control,³⁶ model predictive control,^37,38 and feedback linearization.^39,40 In different application scenarios, these two kinds of methods have their respective advantages.

In fact, in some applications, crane systems exhibit complex flexibility; for instance, when the overhead crane uses a flexible cable for payload transportation, the cable exhibits the characteristics of a flexible cord.⁴¹ Recently, the control of flexible systems on land has attracted much attention in the control and mechatronics society. For example, in He and Ge,⁴¹ cooperative control laws are put forward to limit the flexible cable’s deflection. Based on the dynamic model, a singular perturbation method is applied in Peng et al.⁴² to control the coupling system. Then, in He et al.,⁴³ a hybrid system is obtained and a well-designed controller is proposed to eliminate vibration. Next, in Xie et al.,⁴⁴ a smoother is proposed to reduce the flexible payload’s vibration by command smoothing, which achieves vibration reduction of multi-mode systems. Also, in Fatehi et al.,⁴⁵ two effective controllers are designed on the basis of the singular perturbation method to regulate the crane system with large swing.

The exploitation about marine resources, especially the deep-sea resources, is gradually becoming an increasingly important resource developing trend around the world. As a sequel, deep-sea crane systems are widely applied to exhibit important research values. However, there are few published works on the topic of underwater transportation control. The aforementioned existing methods are designed for flexible systems operating on land without considering underwater situations, which bring much stronger flexibility and are much more challenging. Until now, the research on underwater flexible crane systems is still open. For such systems, the hydrodynamic force will bring larger deflection in the flexible payload under water than on land, that is, producing greater flexibility, causing the payload to vibrate severely, and making the system exhibit much more complicated nonlinear characteristics during transportation. Furthermore, regarding sensors, it is extremely uncertain whether they function well or not in water. Hence, compared with cranes working above water, it is more difficult to achieve precise and efficient transportation for flexible systems in underwater situations. There are some studies about the flexible payload immersed in fluids. For instance, in Shah et al.,⁴⁶ the system model is calculated and a boundary control method is designed to limit the vibration of the rod. Then, in He et al.,⁴⁷ a robust adaptive control method is provided to deal with the vibration of the controlled object in water. In Zhao et al.,⁴⁸ a boundary control method is proposed without simplifying the dynamic equations to restrict the vibration of the flexible riser. To reduce the vibrations in both two directions of the fuel rod, a relatively indirect but comprehensive control law is developed in Shah and Hong⁴⁹ to regulate the movement of the trolley and the bridge. In Shah et al.,⁵⁰ considering the constraint of the rod’s deflection range and a similitude law, a modified input shaping method is developed to limit the flexible rod’s residual vibrations.

In this paper, an anti-vibration polynomial-based trajectory planning method without simplifying and linearizing the dynamic model is proposed. On one hand, the method can restrict the flexible payload’s vibration. On the other hand, it can improve transportation efficiency. Specifically, we divide the velocity of the driving device (e.g. trolley) into three segments: the velocity of the first and last segments is symmetrical, and the velocity of the middle segment is the value allowed by the equipment. Next, the satisfactory performance of the designed method is verified by simulation results. In conclusion, the main contributions of this paper are summarized as follows: (1) without simplifying or linearizing the dynamic model, an anti-vibration polynomial-based trajectory is designed, which can not only meet the requirements of payload transportation but also improve the efficiency; (2) a function $s (t)$ is constructed to obtain the proposed trajectory, which can eliminate the residual vibration; and (3) the proposed method can solve the problem of transporting the flexible payload smoothly in the deep-sea environment, considering complicated hydrodynamics.

The paper is organized as follows. First, in section “Problem statement,” we provide the system model and then summarize the problems to be solved by analyzing the control objective and for the purpose of improving transportation efficiency. Next, in section “Anti-vibration trajectory planning method,” a new planned trajectory, which is designed by planning the trolley’s velocity, is used for this system, and the effectivity of the designed trajectory is verified by numerical simulations. Finally, in section “Conclusion,” we present a summary about the major work of this paper.

Problem statement

Because of the underactuated characteristic of the deep-sea crane system, the vibration at the end point of the payload cannot be directly controlled. In order to ensure accurate and rapid trolley positioning and to suppress the payload’s vibration by constraining the maximum vibration amplitude within a given range, the coupling relationship among state variables must be fully utilized to plan the control input, that is, the acceleration. Furthermore, considering the practice factors, the state variables should meet some specific constraints.

System model

Figure 1 shows a two-dimensional (2D) schematic diagram of the underwater crane system, where in the inertial coordinate $iOj$ , $x (t)$ represents the displacement of the trolley, $γ_{f} (r, t)$ represents the lateral deviation of the payload at r and t, $m_{r}$ is the mass of the payload, and $m_{t}$ is the mass of the trolley. In addition, $u (t)$ denotes the control input of the system and I, E, d, and l denote the area moment of inertia, Young’s modulus, diameter, and length of the payload, respectively.

Figure 1.

The flexible crane system in water.

Let $f_{w} (r, t)$ represent the hydrodynamic force acting on the payload moving in the water. Then, according to Shah et al.,⁵⁰ the expression of $f_{w} (r, t)$ can be given as follows

\begin{matrix} f_{w} (r, t) = \frac{π}{4} ρ_{w} C_{a} (r) d^{2} {\overset{\cdot}{v}}_{r} (r, t) \\ + \frac{1}{2} ρ_{w} C_{d} (r) d v_{r} (r, t) | v_{r} (r, t) | \end{matrix}

(1)

where $v_{r} (r, t) = \overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (r, t)$ , and it is the velocity of the payload. We divide $f_{w} (r, t)$ into two parts as

f_{w} (r, t) = f_{m} (r, t) + f_{d} (r, t)

(2)

where

\begin{matrix} f_{m} (r, t) = \frac{π}{4} ρ_{w} C_{a} (r) d^{2} (\overset{\cdot\cdot}{x} (t) + {\overset{\cdot\cdot}{γ}}_{f} (r, t)) \\ f_{d} (r, t) = \frac{1}{2} ρ_{w} C_{d} (r) d (\overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (r, t)) | (\overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (r, t)) | \end{matrix}

In light of Shah et al.,⁵⁰ the dynamic model of the system is given as follows

(m_{t} + m_{r}) \overset{\cdot\cdot}{x} (t) + m \int_{0}^{l} {\overset{\cdot\cdot}{γ}}_{f} (r, t) dr = u (t)

(3)

EI γ_{f}^{″ ″} (r, t) + m {\overset{\cdot\cdot}{γ}}_{f} (r, t) + c {\overset{\cdot}{γ}}_{f} (r, t) = - m \overset{\cdot\cdot}{x} (t) - f_{d} (r, t)

(4)

where $m = (m_{r} / 4) + (π / 4) ρ_{w} C_{a} (r) d^{2}$ .

Then, we set the following boundary conditions

γ_{f} (0, t) = 0, γ'_{f} (0, t) = 0, {γ ″}_{f} (l, t) = 0, {γ ‴}_{f} (l, t) = 0

(5)

For the sake of distinction, let $γ'_{f} (r, t)$ represent the partial derivative of $γ_{f} (r, t)$ with respect to $r$ . In Table 1, the physical meanings and units of the system parameters are listed in order.

Table 1.

Plant parameters.

Parameters	Physical meaning	Unit
$l$	Payload length	m
$d$	Payload diameter	m
$m_{r}$	Payload mass	kg
$m_{t}$	Trolley mass	kg
$ω_{1}$	Natural frequency	rad/s
$E$	Young’s modulus	GPa
$ρ_{r}$	Density of the payload	kg/m³
$ρ_{w}$	Density of water	kg/m³
$c$	Viscous damping coefficient	N·s/m
$C_{a}$	Added-mass coefficient	–
$C_{d}$	Drag coefficient	–

Control objective

Considering the actual operating environment of the deep-sea crane system, the control objective can be described as follows:

The time to complete the entire transportation process is defined as $T$ . When $t = T$ , the trolley should arrive at the target point $p_{d}$ from the initial position, and the payload vibration should converge to the equilibrium point without obvious oscillation, that is, the velocity, acceleration, and jerk of the trolley, as well as the vibration velocity, vibration acceleration, and vibration jerk of the payload, all should be zero at $T$ and is given by

\begin{matrix} x (0) = 0, \overset{\cdot}{x} (0) = 0, \overset{\cdot\cdot}{x} (0) = 0, \overset{\dots}{x} (0) = 0, \\ x (T) = p_{d}, \overset{\cdot}{x} (T) = 0, \overset{\cdot\cdot}{x} (T) = 0, \overset{\dots}{x} (T) = 0, \\ γ_{f} (r, 0) = 0, {\overset{\cdot}{γ}}_{f} (r, 0) = 0, \\ {\overset{\cdot\cdot}{γ}}_{f} (r, 0) = 0, {\overset{\dots}{γ}}_{f} (r, 0) = 0, \\ γ_{f} (r, T) = 0, {\overset{\cdot}{γ}}_{f} (r, T) = 0, \\ {\overset{\cdot\cdot}{γ}}_{f} (r, T) = 0, {\overset{\dots}{γ}}_{f} (r, T) = 0 \end{matrix}

(6)

Based on hardware restrictions and safety concerns, the state variables, that is, trolley velocity, payload deviation, and so on, should be reasonably limited within a suitable range during the entire transportation process. Therefore, the variables should satisfy the following constraints

\begin{matrix} | \overset{\cdot}{x} (t) | \leq v_{\max}, | \overset{\cdot\cdot}{x} (t) | \leq a_{\max}, \\ | \overset{\dots}{x} (t) | \leq j_{\max}, | γ_{f} (r, t) | \leq γ_{f \max}, \\ | {\overset{\cdot}{γ}}_{f} (r, t) | \leq v_{γ_{f \max}}, | {\overset{\cdot\cdot}{γ}}_{f} (r, t) | \leq a_{γ_{f \max}} \end{matrix}

(7)

where $v_{\max}, a_{\max}, j_{\max}, γ_{f \max}, v_{γ_{f \max}}, and a_{γ_{f \max}}$ denote the maximum values within the allowed range of the corresponding variables.

The optimal time under certain conditions

To improve the efficiency of the system, the value of $T$ is expected to be as small as possible, on the premise of satisfying the listed conditions. Taking all the constraints into consideration, we can formulate and solve the following problem to obtain the most appropriate time

\begin{matrix} \min & T, \\ s . t . & | \overset{\cdot}{x} (t) | \leq v_{\max}, | \overset{\cdot\cdot}{x} (t) | \leq a_{\max}, \\ | \overset{\dots}{x} (t) | \leq j_{\max}, | γ_{f} (r, t) | \leq γ_{f \max}, \\ | {\overset{\cdot}{γ}}_{f} (r, t) | \leq v_{γ_{f \max}}, | {\overset{\cdot\cdot}{γ}}_{f} (r, t) | \leq a_{γ_{f \max}} \end{matrix}

(8)

Anti-vibration trajectory planning method

For the purpose of simplifying the calculation, only the deviation value of the end point of the flexible payload is considered. According to the practical situation, if the vibration of the end point of the payload meets the requirements, the vibration of the other points must also be within the allowable range. Hence, this paper only considers the state of the end point, which can equivalently capture the state of the entire payload. At this point, $r$ can be replaced by $l$ , and then the model coupling relationship given by equation (4) is rewritten as follows

\begin{matrix} c {\overset{\cdot}{γ}}_{f} (t) = - (\frac{1}{2}) ρ_{w} C_{d} (r) d (\overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (t)) | \overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (t) | \\ - m {\overset{\cdot\cdot}{γ}}_{f} (t) - m \overset{\cdot\cdot}{x} (t) \end{matrix}

(9)

From equation (9), we construct a signal

s (t) = x (t) + γ_{f} (t)

(10)

By substituting equation (10) into equation (9), one has that

{\overset{\cdot}{γ}}_{f} (t) = - \frac{1}{c} (m \overset{\cdot\cdot}{s} (t) + (\frac{1}{2}) ρ_{w} C_{d} (r) d \overset{\cdot}{s} (t) | \overset{\cdot}{s} (t) |)

(11)

\overset{\cdot}{x} (t) = \overset{\cdot}{s} (t) + \frac{1}{c} (m \overset{\cdot\cdot}{s} (t) + (\frac{1}{2}) ρ_{w} C_{d} (r) d \overset{\cdot}{s} (t) | \overset{\cdot}{s} (t) |)

(12)

After that, all state variables can be represented by the signal $s (t)$ . Then, trajectory planning can be transformed into the problem of designing $s (t)$ . In the subsequence, we can define $s (t)$ as a polynomial expression.

Trajectory planning

If the velocity of the trolley can reach the maximum value within a limited time and last for a while, the running time will be further shortened and the transportation efficiency will be improved compared with the single-peak velocity trajectory, which is a trapezoid velocity trajectory containing three curve segments (increasing velocity, constant velocity, and decreasing velocity). The maximum velocity of the trolley can be obtained based on the limit range of the deflection, that is, it is feasible to plan the trajectory of the trolley’s velocity.

To realize the control objective, the velocity is divided into three segments, and the acceleration can be guaranteed to be continuous. The first segment is designed as a polynomial from $t = 0$ to $t = t_{1}$ , the third segment is also a polynomial from $t = t_{2}$ to $t = T$ , and the second segment is a constant $v_{\max}$ from $t = t_{1}$ to $t = t_{2}$ . The derivative of $s (t)$ with respect to $t$ is expressed as follows

\overset{\cdot}{s} (t) = \overset{\cdot}{x} (t) + {\overset{\cdot}{γ}}_{f} (t)

First, to obtain the velocity of the first segment, $\overset{\cdot}{s} (t)$ is designed as follows

\overset{\cdot}{s} (t) = p_{0} + (\sum_{i = 1}^{6} p_{i} η^{i})

(13)

where $η = t / t_{1}$ , $0 \leq t \leq t_{1}$ .

The partial equality constraints of $\overset{\cdot}{s} (t)$ can be derived based on the constraints of $x (t)$ and $γ_{f} (t)$ , which can be expressed as

\begin{matrix} \overset{\cdot}{s} (0) = 0, \overset{\cdot\cdot}{s} (0) = 0, \overset{\dots}{s} (0) = 0, \\ \overset{\cdot}{s} (t_{1}) = v_{\max}, \overset{\cdot\cdot}{s} (t_{1}) = 0, \overset{\dots}{s} (t_{1}) = 0 \end{matrix}

(14)

By substituting equation (14) into equation (13), the following equations and coefficient relationships can be obtained

{\begin{matrix} p_{0} = p_{1} = p_{2} = 0 \\ p_{3} = - p + 10 \\ p_{4} = 3 p - 15 \\ p_{5} = - 3 p + 6 \\ p_{6} = p \end{matrix}

(15)

After that, we let

\begin{matrix} \overset{\cdot}{x} (t) = g_{3} η^{3} + g_{4} η^{4} + g_{5} η^{5} + g_{6} η^{6}, \\ {\overset{\cdot}{γ}}_{f} (t) = h_{3} η^{3} + h_{4} η^{4} + h_{5} η^{5} + h_{6} η^{6} \end{matrix}

(16)

Then, the constraints of $\overset{\cdot}{x} (t)$ and ${\overset{\cdot}{γ}}_{f} (t)$ are written as

\begin{matrix} \overset{\cdot}{x} (0) = 0, \overset{\cdot\cdot}{x} (0) = 0, \overset{\dots}{x} (0) = 0, \\ \overset{\cdot}{x} (t_{1}) = v_{\max}, \overset{\cdot\cdot}{x} (t_{1}) = 0, \overset{\dots}{x} (t_{1}) = 0, \\ {\overset{\cdot}{γ}}_{f} (0) = 0, {\overset{\cdot\cdot}{γ}}_{f} (0) = 0, {\overset{\dots}{γ}}_{f} (0) = 0, \\ {\overset{\cdot}{γ}}_{f} (t_{1}) = 0, {\overset{\cdot\cdot}{γ}}_{f} (t_{1}) = 0, {\overset{\dots}{γ}}_{f} (t_{1}) = 0 \end{matrix}

(17)

By substituting equation (17) into equation (16), the coefficient relationships can be rewritten as follows

{\begin{matrix} g_{3} = - g + 10 v_{m} \\ g_{4} = 3 g - 15 v_{m} \\ g_{5} = - 3 g + 6 v_{m} \\ g_{6} = g \end{matrix}

(18)

{\begin{matrix} h_{3} = - h \\ h_{4} = 3 h \\ h_{5} = - 3 h \\ h_{6} = h \end{matrix}

(19)

Actually, if the initial values of any two of the three parameters are given, we can then obtain the value of the last parameter. Specifically, from equations (10), (13), (15), (16), (18), and (19), we can derive that $g + h = p$ ; that is to say, if the two parameter values are given, we can obtain the remaining one. In addition, according to the physical significance of $s (t)$ , it is the sum of the trolley displacement and the payload end-point displacement, indicating that $\overset{\cdot}{x} (t) \geq 0$ , $\overset{\cdot}{s} (t) \geq 0$ , and $s (t) \leq x (t)$ . Then by trial and error, we can adjust the parameters and determine suitable values.

So far, the first velocity expression containing an unknown parameter $t_{1}$ is obtained. Nevertheless, it is necessary to acquire the optimal $t_{1}$ to obtain the first trajectory. It is clear that the transportation efficiency will be improved by reducing $t_{1}$ on the premise of ensuring tracking performances.

To do so, we use the dichotomy to solve the problem in equation (8) and obtain the optimal time. The pseudocode is given in Algorithm 1, where $T_{s}$ and $T_{b} \in R^{+}$ , respectively, represent the minimum and maximum values of $T$ ; $μ$ denotes the allowable error, which is used as the ending condition of the program and is set artificially as needed; $v_{\max}, a_{\max}, j_{\max}, γ_{f \max}, v_{γ_{f \max}}, and a_{γ_{f \max}}$ are the upper limits of the constraints in equation (7); and $T^{*}$ is the optimal value of $T$ .

Algorithm 1: Algorithm for solving optimal time problem in equation (8)
Input: $T_{s}, T_{b}, v_{\max}, a_{\max}, j_{\max}, γ_{f \max}, v_{γ_{f \max}}, a_{γ_{f \max}}$
Output: $T^{*}$
1: while $\| T_{b} - T_{s} \| > μ$ do
2: let $T_{t} = (T_{s} + T_{b}) / 2$
3: if the inequality constraints in equation (7) are all satisfied then
4: let $T_{b} = T_{t}$
5: else
6: let $T_{s} = T_{t}$
7: end if
8: end while
9: let $T^{*} = T_{t}$

Based on equation (7), $t_{1}$ can be solved by Algorithm 1. Thus, the velocity trajectory of the first segment is planned. The velocity of the trolley in the second segment can be derived directly and is constant, that is, $\overset{\cdot}{x} (t) = v_{\max} (t_{1} \leq t \leq t_{2})$ , where $t_{2}$ is an unknown parameter, which can be obtained by the third segment. In this paper, the terminal velocity and the first segment are symmetrical, the time relationship $T - t_{2} = t_{1}$ is satisfied, and the displacement relationship $p_{d} - v_{\max} (t_{2} - t_{1}) = 2 x (t_{1})$ is also satisfied. Since the first expression is known, the final velocity expression can be obtained after mathematical arrangements, and $t_{2}$ and $T$ can also be calculated. Then, the entire trajectory planning process is completed.

Numerical simulation results

To obtain the deflection values of the flexible payload based on the trolley movement, the finite difference method (FDM) is applied to solve the partial differential equation (PDE) in equation (4). For the simulations, the parameters are set as in Table 2, which are from Shah et al.⁵⁰

Table 2.

System parameters.

Parameters	Value	Unit
$l$	1.0	m
$d$	0.008	m
$m_{r}$	0.0369	kg
$m_{t}$	3.8011	kg
$ω_{1}$	9.52	rad/s
$E$	38.9	GPa
$ρ_{r}$	1204.5	kg/m³
$ρ_{w}$	1000.0	kg/m³
$c$	0.6	N·s/m
$C_{a}$	0.93	–
$C_{d}$	1.28	–

Figure 2 indicates the simulation results of the planned anti-vibration trajectory with $p_{d} = 1.2 m$ , $v_{\max} = 0.35 m / s$ , $p = 10$ , and $g = 3$ . It indicates that the proposed trajectory can guarantee the continuity of the trolley acceleration.

Figure 2.

Displacement, velocity, and acceleration trajectories of the trolley.

For the purpose of verifying the effectiveness of the planned trajectory, the paper compares the trajectory planning method with the existing input shaping method in Shah et al.⁵⁰ The corresponding simulation results are shown in Figure 3. Compared with the input shaping method, what we can see from Figure 3 is that the transportation time of the anti-vibration trajectory planning method is shorter. Therefore, in terms of transportation efficiency, the proposed trajectory planning method is better than the input shaping method.

Figure 3.

Results of the proposed and comparative methods.

Besides, for the input shaping method, it is noticed from Figure 3 that the trolley acceleration, which is the system’s control input, will suddenly change at some point, that is, the driving force will change suddenly. By contrast, this phenomenon is avoided by the proposed method, which is smoother. Moreover, compared with existing methods (e.g. input shaping method), the proposed method ensures that the accelerations are continuous and also protects the actuators in practice.

Figure 4 is a three-dimensional (3D) diagram of the deflection of each position of the flexible payload with respect to time using the anti-vibration trajectory planning method, and it reflects the vibration response of the entire flexible payload to the velocity command.

Figure 4.

Three-dimensional diagram of the anti-vibration trajectory planning method.

Conclusion

In this paper, with no linearization to the dynamic model, a new anti-vibration polynomial-based trajectory planning method is proposed, which considers the constraints of all state variables while achieving the control objective. The design of the anti-vibration trajectory is to divide the velocity into three segments and make the trolley move at the maximum velocity for longer time, which improves the transportation efficiency greatly. In the end, simulation results verify the performance of the proposed method. In the near future, we will implement hardware experiments for further verification, and other types of trajectories will also be considered apart from polynomials.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by the National Natural Science Foundation of China (Grant Nos U1706228 and 61873134) and in part by the open project of Industrial Robot Application of Fujian University Engineering Research Center, Minjiang University (Grant No. MJUKF-IRA201801)

ORCID iDs

Yue Wang

Ning Sun

References

Zhang

, et al. Modeling and energy-based sway reduction control for tower crane systems with double-pendulum and spherical-pendulum effects. Meas Control 2020; 53: 141–150.

Abdel-Rahman

Nayfeh

Masoud

ZN.

Dynamics and control of cranes: a review. Modal Anal 2003; 9(7): 863–908.

Sun

Lyu

Fang

, et al. Robust LQR anti-swing control for quay-side crane system with variable load. In: 2018 IEEE 8th annual international conference on CYBER technology in automation, control, and intelligent systems (CYBER), Tianjin, China, 19–23 July 2018, pp. 796–801. New York: IEEE.

Ouyang

Zhang

. Energy-shaping-based nonlinear controller design for rotary cranes with double-pendulum effect considering actuator saturation. Automat Constr 2020; 111: 103054.

Lee

. 3D cooperative control of tower cranes using robust adaptive techniques. J Frankl Inst 2017; 354(18): 8333–8357.

Sun

Fang

Chen

, et al. Nonlinear stabilizing control for ship-mounted cranes with ship roll and heave movements: design, analysis, and experiments. IEEE T Syst Man Cy: S 2017; 48(10): 1781–1793.

Ramli

Mohamed

Abdullahi

, et al. Control strategies for crane systems: a comprehensive review. Mech Syst Signal Pr 2017; 95: 1–23

Yang

Sun

Chen

, et al. Neural network-based adaptive antiswing control of an underactuated ship-mounted crane with roll motions and input dead zones. IEEE T Neur Net Lear 2020; 31: 901–914.

Tuan

Cuong

Lee

, et al. Nonlinear feedback control of container crane mounted on elastic foundation with the flexibility of suspended cable. J Vib Control 2016; 22(13): 3067–3078.

10.

Cheng

Chen

. Controller design for an overhead crane system with uncertainty. Control Eng Pract 1996; 4(5): 645–653.

11.

Ouyang

Zhang

. Tracking and load sway reduction for double-pendulum rotary cranes using adaptive nonlinear control approach. Int J Robust Nonlin 2020; 30: 1872–1885.

12.

Zhang

Rong

, et al. Nonlinear coupling control method for underactuated three-dimensional overhead crane systems under initial input constraints. T I Meas Control 2018; 40(2): 413–424.

13.

Sun

Chen

, et al. Antiswing cargo transportation of underactuated tower crane systems by a nonlinear controller embedded with an integral term. IEEE T Autom Sci Eng 2019; 16(3): 1387–1398.

14.

Zhang

Ouyang

, et al. Adaptive integral sliding mode control with payload sway reduction for 4-DOF tower crane systems. Nonlinear Dynam. Epub ahead of print 19 January 2020. DOI: 10.1007/s11071-020-05471-3.

15.

Huang

Gao

Ralescu

, et al. Adaptive hierarchical sliding mode control based on fuzzy neural network for an underactuated system. Adv Mech Eng 2018; 10(9). DOI: 10.1177/1687814018799554.

16.

Zhang

. Finite-time model-free trajectory tracking control for overhead cranes subject to model uncertainties, parameter variations and external disturbances. T I Meas Control 2019; 41: 3516–3525.

17.

Gao

Zhang

Chen

, et al. Energy-based control design of an underactuated 2-dimensional TORA system. In: Proceedings of the IEEE/RSJ international conference on intelligent robots and systems, St. Louis, MO, 10–15 October 2009, pp. 1296–1301. New York: IEEE.

18.

Sun

Xia

, et al. Adaptive tracking control of wheeled inverted pendulums with periodic disturbances. IEEE T Cybernetics. Epub ahead of print 20 December 2018. DOI: 10.1109/TCYB.2018.2884707.

19.

Bessa

Otto

Kreuzer

, et al. An adaptive fuzzy sliding mode controller for uncertain underactuated mechanical systems. J Vib Control 2019; 25(9): 1521–1535.

20.

Sun

Xia

, et al. Adaptive fuzzy tracking control of flexible-joint robots with full-state constraints. IEEE T Syst Man Cy: S 2018; 49(11): 2201–2209.

21.

Chen

Sun

. Nonlinear control of underactuated systems subject to both actuated and unactuated state constraints with experimental verification. IEEE T Ind Electron. Epub ahead of print 15 October 2019. DOI: 10.1109/TIE.2019.2946541.

22.

Sun

Yang

, et al. Nonlinear motion control of complicated dual rotary crane systems without velocity feedback: design, analysis, and hardware experiments. IEEE T Autom Sci Eng. 2020; 17(2): 1017–1029. DOI: 10.1109/TASE.2019.2961258.

23.

Gao

Bao

Xie

, et al. Passivity-based control of two-dimensional translational oscillator with rotational actuator. T I Meas Control 2014; 36(1): 111–118.

24.

. Adaptive control of the TORA system with partial state constraint. T I Meas Control 2019; 41(4): 1172–1177.

25.

Sun

Fang

, et al. Nonlinear continuous global stabilization control for underactuated RTAC systems: design, analysis, and experimentation. IEEE/ASME T Mech 2017; 22(2): 1104–1115.

26.

Lai

Wang

, et al. Stable control strategy for planar three-link underactuated mechanical system. IEEE/ASME T Mech 2016; 21(3): 1345–1356.

27.

Sun

Liang

, et al. Adaptive control for pneumatic artificial muscle systems with parametric uncertainties and unidirectional input constraints. IEEE T Ind Inform 2020; 16(2): 969–979.

28.

Prattichizzo

Malvezzi

Gabiccini

, et al. On the manipulability ellipsoids of underactuated robotic hands with compliance. Robot Auton Syst 2012; 60(3): 337–346.

29.

Sun

Chen

, et al. An energy-optimal solution for transportation control of cranes with double pendulum dynamics: design and experiments. Mech Syst Signal Pr 2018; 102: 87–101

30.

Hong

Huh

Hong

. Command shaping control for limiting the transient sway angle of crane systems. Int J Control Autom 2003; 1(1): 43–53.

31.

Mar

Goyal

Nguyen

, et al. Combined input shaping and feedback control for double-pendulum systems. Mech Syst Signal Pr 2017; 85: 267–277

32.

Seifried

. Two approaches for feedforward control and optimal design of underactuated multibody systems. Multibody Syst Dyn 2012; 27(1): 75–93.

33.

Ngo

Nguyen

, et al. Fuzzy sliding mode control of an offshore container crane. Ocean Eng 2017; 140: 125–134.

34.

Chwa

. Sliding-mode-control-based robust finite-time antisway tracking control of 3-D overhead cranes. IEEE T Ind Electron 2017; 64(8): 6775–6784

35.

Qian

Ding

Lee

, et al. Suppression of chaotic behaviors in a complex biological system by disturbance observer-based derivative-integral terminal sliding mode. IEEE/CAA J Automatica Sinica 2019; 7(1): 126–135.

36.

Smoczek

. Fuzzy crane control with sensorless payload deflection feedback for vibration reduction. Mech Syst Signal Pr 2014; 46(1): 70–81.

37.

Xia

Zhu

. Model predictive control for improving operational efficiency of overhead cranes. Nonlinear Dynam 2015; 79(4): 2639–2657

38.

, et al. Speed sensorless model predictive control method for a direct-drive wind energy conversion system. Meas Control 2019; 52: 1394–1402.

39.

Park

Chwa

Hong

. A feedback linearization control of container cranes: varying rope length. Int J Control Autom 2007; 5(4): 379–387.

40.

Lee

Kim

, et al. A study on load position control and vibration attenuation in crane operation using sub-actuator. Meas Control 2019; 52: 794–803.

41.

. Cooperative control of a nonuniform gantry crane with constrained tension. Automatica 2016; 66: 146–154.

42.

Peng

Liu

. Boundary control for a flexible inverted pendulum system based on a PDE model. Asian J Control 2018; 20(1): 12–21.

43.

Zhang

. Adaptive control of a flexible crane system with the boundary output constraint. IEEE T Ind Electron 2013; 61(8): 4126–4133

44.

Xie

Huang

Liang

. Vibration reduction for flexible systems by command smoothing. Mech Syst Signal Pr 2013; 39(1–2): 461–470.

45.

Fatehi

Eghtesad

Necsulescu

, et al. Tracking control design for a multi-degree underactuated flexible-cable overhead crane system with large swing angle based on singular perturbation method and an energy-shaping technique. J Vib Control 2019; 25(11): 1752–1767.

46.

Shah

Piao

Kim

, et al. Active vibration control of a flexible rod in water. In: Proceedings of the 17th international conference on control, automation and systems (ICCAS), Jeju, South Korea, 18–21 October 2017, pp. 1018–1023. New York: IEEE.

47.

How

BVE

, et al. Robust adaptive boundary control of a flexible marine riser with vessel dynamics. Automatica 2011; 47(4): 722–732.

48.

Zhao

Liu

, et al. Control design for a vibrating flexible marine riser system. J Frankl Inst 2017; 354(18): 8117–8133.

49.

Shah

Hong

. Active vortex-induced vibration control of a flexible rod transported in water. In: Proceedings of the 11th Asian control conference (ASCC), Gold Coast, QLD, Australia, 17–20 December 2017, pp. 2429–2434. New York: IEEE.

50.

Shah

Hong

Choi

. Open-loop vibration control of an underwater system: application to refueling machine. IEEE/ASME T Mech 2017; 22(4): 1622–1632.

Point-to-point motion control for flexible crane systems working in the deep sea

Abstract

Keywords

Introduction

Problem statement

System model

Control objective

The optimal time under certain conditions

Anti-vibration trajectory planning method

Trajectory planning

Numerical simulation results

Conclusion

Footnotes

Declaration of conflicting interests

Funding

ORCID iDs

References