Nonlinear stability control method and system for offshore crane based on neural network

Abstract

The invention discloses a nonlinear stability control method and system for an offshore crane based on a neural network. The nonlinear stability control method comprises the following steps: S1, acquiring system parameters of a marine crane; s2, constructing an energy function; s3, constructing a radial basis function neural network to estimate lumped unknown dynamics in the system; s4, constructing a nonlinear stability controller in combination with the radial basis function neural network; and S5, the designed controller is used for controlling, and positioning and swing eliminating of the ship crane system are achieved. The method comprises the following steps: establishing a ship crane model by utilizing a Lagrange modeling method, constructing a Lyapunov energy function containing potential energy and kinetic energy based on the model, then deducing a basic form of an adaptive controller based on a first-order derivative of the energy function, and estimating an unknown dynamic state of a system by utilizing a designed radial basis function neural network. The designed controller can realize the positioning of the cantilever and the load, and has a better inhibition effect on the shaking of the load.

Description

Technical Field

[0001] The invention relates to the technical field of anti-sway motion control of bridge cranes, and in particular to a nonlinear stability control method and system for offshore cranes based on neural networks. Background Art

[0002] Floating crane is a kind of engineering vessel specially used for lifting operation on water. It has a wide range of applications, such as using it to complete cargo loading and unloading in ports; completing salvage and rescue in maritime accidents; completing port construction, bridge building, and equipment installation in various harbors and offshore marine engineering projects; completing equipment outfitting and repair in shipyards; in recent years, with the development of offshore oil, relying on super-large floating cranes, completing the construction and installation of drilling platforms.

[0003] In the prior art, research on the complex dynamics of marine cranes mainly focuses on modeling and control system design, but there are still many difficulties in practical applications. Complex external disturbances in the marine environment, such as unknown friction and random load changes caused by sea breeze, are usually not accurately described by traditional physical modeling methods. In addition, the nonlinear characteristics of the hull and the crane during operation, as well as the double pendulum effect of the cantilever, hook and load, make the system highly coupled and dynamically uncertain. The existence of these unknown dynamics significantly increases the difficulty of controller design, making it difficult to ensure control accuracy and system robustness. Traditional control systems usually rely on precise system parameters, but in actual situations, due to the limitations of sensors, it is difficult to directly obtain a series of time-varying unknown parameters such as mechanical friction and wind force, and uncertain system dynamics will inevitably lead to a decrease in the positioning accuracy of the crane load, further reducing the reliability of the system. Summary of the invention

[0004] The purpose of the present invention is to provide a nonlinear stability control method and system for an offshore crane based on a neural network, so as to solve the complex external disturbances in the marine environment proposed in the above-mentioned background technology, such as unknown friction, random load changes caused by sea breeze, etc., which are usually unable to be accurately described by traditional physical modeling methods. In addition, the nonlinear characteristics of the hull and the crane during operation, as well as the double pendulum effect of the cantilever, hook and load, make the system highly coupled and dynamically uncertain. The existence of these unknown dynamics significantly increases the difficulty of controller design, resulting in the problem that control accuracy and system robustness are difficult to guarantee.

[0005] To achieve the above object, the present invention provides the following technical solution: a nonlinear stability control method for an offshore crane based on a neural network, comprising the following steps:

[0006] S1. Obtain the parameters of the ship crane system and establish a three-dimensional ship crane system dynamic model considering the roll and pitch motion of the ship;

[0007] S2, constructing an energy function to describe the energy change according to the total energy change rate of the marine crane system;

[0008] S3, constructing a radial basis neural network to estimate the aggregate unknown dynamics in the system;

[0009] S4, construct a nonlinear stable controller by combining radial basis neural network;

[0010] S5. Use the designed controller to control and realize the positioning and anti-sway of the marine crane system.

[0011] Preferably, in step S1, the Lagrangian modeling equation is used to construct a dynamic model of the marine crane system in combination with wind resistance and wave disturbance existing in practical applications.

[0012] Preferably, the dynamic model of the marine crane system is expressed as:

[0013]

[0014] M=[m ij ] 6×6 ,C=[c ij ] 6×6 ,G=[g i ] 6×1 ;

[0015] U=[0 0 0 0 τ5+Δτ5 τ6+Δτ6] T ;

[0016]

[0017] Where: M(ζ) is the mass matrix, is the centripetal-Coriolis force matrix, G(ζ) is the gravity vector, U is the input vector, F is the wind resistance and friction vector, ζ is the state variable, and are the velocity and acceleration of the state variables, τ5 and τ6 are the heave and rotation moments of the cantilever, Δτ5 and Δτ6 represent the nonlinear terms in the input dead zone, Represents the drag coefficient, f r5 and f r6 is the mechanical friction of the cantilever, f a ,δ5,f b and δ6 represents the static friction coefficient, κ a and κ b Represents the sliding friction coefficient.

[0018] Preferably, the internal elements of the matrix and vector have the following expressions:

[0019]

[0020] m 13 =m2l1l2(1+ζ1ζ3),m 14 =m2l1l2ζ1ζ4,

[0021]

[0022] m 21 =m 12 , m 23 =m2l1l2ζ2ζ3;

[0023] m 24 =m2l1l2(1+ζ2ζ4),m 25 =-(m1+m2)l1Lζ2S5;

[0024]

[0025] m 31 =m 13 ,m 32 =m 23 ,

[0026] m 35 =m2l2L(C5-ζ3S5),

[0027] m 41 =m 14 ,m 42 =m 24 ,m 43 =m 34 ,

[0028] m 45 =-m2l2Lζ4S5,

[0029] m 51 =m 15 ,m 52 =m 25 ,m 53 =m 35 ,m 55 =(m1+m2)L 2 +I y ;

[0030] m 56 =-((m1+m2)l1Lζ2C5+m2l2Lζ4C5),m 54 =m 45 ;

[0031]

[0032] C 11 =0,

[0033]

[0034] C 22 =0,C 24 =0;

[0035]

[0036] C 31 =0, C 33 =0;

[0038] C 41 =m2l1l2ζ4ζ1;

[0040] C 42 =0,C 44 =0, C 51 =0,C 52 =0,C 53 =0,C 54 =0,C 55 =0; C 61 =0;

[0042] C 62 =0,C 63 =0,

[0043] g1

[0044] =(m1+m2)l1gζ1;

[0045] g4=m2l2gζ4, g2=(m1+m2)l1gζ2, g3=m2l2gζ3, g6=0;

[0046] g5=-gS5(0.5M0L+(m1+m2)L-0.5M1L1);

[0047] Where M0 and M1 are the masses of the ballast and boom, m0 and m1 are the masses of the hook and load, L, L1 and l1 are the lengths of the boom, ballast and rope, l2 is the distance between the hook and the load, Ix ,I y ,I z and I b are all parameters related to the moment of inertia, g is the acceleration due to gravity, C (*) and S (*) are the abbreviations of cos(*) and sin(*), τ5 and τ6 are the torques driving the boom to rise and fall and rotate, ζ1, ζ2, ζ3 and ζ4 are the swing angles of the hook and the load, ζ5 and ζ6 are the heave angle and rotation angle of the boom, respectively. i (i=1…6) are the corresponding angles in the inertial coordinate system. Their corresponding relationship with the angles in the crane non-relational coordinate system is as follows:

[0048] ζ1=eta1-d1, ζ2=eta2-d2, ζ3=eta3-d1;

[0049] ζ4=eta4-d2, ζ5=eta5+d1, ζ6=eta6+d2;

[0050] where η i (i=1…6) is the corresponding state quantity in the hull coordinate system, d1 and d2 are coupled disturbance variables, which are formed by the roll, pitch and system state quantity coupling of the hull. The specific expression is:

[0051]

[0052] Preferably, in step S2, the total energy function of the crane system is:

[0053]

[0054] Taking the derivative, we get:

[0055]

[0056] in,

[0057] e=[ζ1 ζ2 ζ3 ζ4 e5 e6] T ;

[0058] e5 and e6 are the tracking errors of the cantilever heave angle and rotation angle, respectively, that is, e5 = ζ5-ζ 5r , e6=ζ6-ζ 6r .

[0059] Preferably, in step S3, the unknown dynamic expression of the crane system is:

[0060]

[0061] Among them, 11 and21 represents an unknown dynamic with an initial value of zero, and Υ 12 and 22 represents the unknown dynamics that are not initially zero, Δτ 51 and Δτ 61 represents the items in Δτ5 and Δτ6 related to the system state quantity, while Δτ 52 and Δτ 62 Represents the items in Δτ5 and Δτ6 related to the mechanical structure of the system, and the expression of the radial basis neural network is:

[0062]

[0063]

[0064] Preferably, in step S3, the update law expression of the neural network is:

[0065]

[0066] where Ψ1 and Ψ2 are positive definite diagonal matrices.

[0067] Preferably, in step S4, the nonlinear stability controller expression is:

[0068]

[0069] Among them, δ π =2π / (ξ 6M -ξ 6m ) 2 , k p5 , k p6 , k d5 , k d6 , k s5 and k s6 All are control parameters and meet the following selection rules:

[0070]

[0071] A nonlinear stability control system for an offshore crane based on a neural network, comprising a model building module, an energy analysis module, a radial basis neural network module, a controller design module and a nonlinear stability control module;

[0072] The model building module is configured to obtain the parameters of the ship crane system, consider the roll and pitch motion of the ship hull, and establish a dynamic model of the ship crane system;

[0073] The energy analysis module is configured to construct an energy function to describe energy changes according to the total energy change rate of the ship crane system;

[0074] The radial basis neural network module is configured to design an auxiliary function that can replace the position parameters of the control system;

[0075] The controller design module is configured to design a nonlinear stable controller according to the radial basis neural network and the system energy function analysis results;

[0076] The nonlinear stability control module is configured to use a nonlinear stability controller to perform control to achieve positioning and sway elimination of the marine crane system.

[0077] Compared with the prior art, the present invention has the following beneficial effects:

[0078] The present invention uses the Lagrangian modeling method to establish a ship crane model, and constructs a Lyapunov energy function containing potential energy and kinetic energy based on the model. Subsequently, based on the first-order derivative of the energy function, the basic form of the adaptive controller is derived, and the designed radial basis function neural network is used to estimate the unknown dynamics of the system. The designed controller can realize the positioning of the cantilever and the load, and has a good suppression effect on the sway of the load. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 A control method flow chart of a nonlinear stability control method and system of an offshore crane based on a neural network according to the present invention;

[0080] Figure 2 It is a principle diagram of a nonlinear stable control method and system of an offshore crane based on a neural network according to the present invention;

[0081] Figure 3 This is a graph showing the experimental results of a PID controller of a nonlinear stable control method and system for an offshore crane based on a neural network according to the present invention;

[0082] Figure 4 This is a graph of experimental results of a neural network-based offshore crane nonlinear stability control method and a system PID controller under different system parameters. DETAILED DESCRIPTION

[0083] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the implementation regulations described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0084] Example 1: Reference Figure 1-Figure 4 As shown: A nonlinear stability control method for offshore crane based on neural network, comprising the following steps:

[0085] Step 1: Obtain the parameters of the marine crane system, combine the wind resistance and wave disturbance existing in actual applications, and use the Lagrangian modeling equation to build a dynamic model of the marine crane system;

[0086] The dynamic model of the marine crane system is expressed as:

[0087]

[0088] M=[m ij ] 6×6 ,C=[c ij ] 6×6 ,G=[g i ] 6×1 ;

[0089] U=[0 0 0 0 τ5+Δτ5 τ6+Δτ6] T ;

[0090]

[0091] Where: M(ζ) is the mass matrix, is the centripetal-Coriolis force matrix, G(ζ) is the gravity vector, U is the input vector, F is the wind resistance and friction vector, ζ is the state variable, and are the velocity and acceleration of the state variables, τ5 and τ6 are the heave and rotation moments of the cantilever, Δτ5 and Δτ6 represent the nonlinear terms in the input dead zone, Represents the drag coefficient, f r5 and f r6 is the mechanical friction of the cantilever, f a ,δ5,f b and δ6 represents the static friction coefficient, κ a and κ b represents the sliding friction coefficient, and the internal elements of the above matrices and vectors have the following expressions:

[0092]

[0093] m 13 =m2l1l2(1+ζ1ζ3),m 14 =m2l1l2ζ1ζ4,

[0094]

[0095] m 21 =m 12 , m 23=m2l1l2ζ2ζ3;

[0096] m 24 =m2l1l2(1+ζ2ζ4),m 25 =-(m1+m2)l1Lζ2S5;

[0097]

[0098] m 31 =m 13 ,m 32 =m 23 ,

[0099] m 35 =m2l2L(C5-ζ3S5),

[0100] m 41 =m 14 ,m 42 =m 24 ,m 43 =m 34 ,

[0101] m 45 =-m2l2Lζ4S5,

[0102] m 51 =m 15 ,m 52 =m 25 ,m 53 =m 35 ,m 55 =(m1+m2)L 2 +I y ;

[0103] m 56 =-((m1+m2)l1Lζ2C5+m2l2Lζ4C5),m 54 =m 45 ;

[0104]

[0105] C 11 =0,

[0106]

[0107] C 22 =0,C 24 =0;

[0108]

[0109] C 31 =0, C 33 =0;

[0111] C 41 =m2l1l2ζ4ζ1;

[0113] C 42 =0,C 44 =0, C 51 =0,C 52 =0,C 53 =0,C 54 =0,C 55 =0;

[0114] C 61 =0;

[0116] C 62 =0,C 63 =0,

[0117] g1

[0118] =(m1+m2)l1gζ1;

[0119] g4=m2l2gζ4, g2=(m1+m2)l1gζ2, g3=m2l2gζ3, g6=0;

[0120] g5=-gS5(0.5M0L+(m1+m2)L-0.5M1L1);

[0121] Where M0 and M1 are the masses of the ballast and boom, m0 and m1 are the masses of the hook and load, L, L1 and l1 are the lengths of the boom, ballast and rope, l2 is the distance between the hook and the load, I x ,I y ,I z and I b are all parameters related to the moment of inertia, g is the acceleration due to gravity, C (*) and S (*) are the abbreviations of cos(*) and sin(*), τ5 and τ6 are the torques driving the boom to rise and fall and rotate, ζ1, ζ2, ζ3 and ζ4 are the swing angles of the hook and the load, ζ5 and ζ6 are the heave angle and rotation angle of the boom, respectively. i(i=1…6) are the corresponding angles in the inertial coordinate system. Their corresponding relationship with the angles in the crane non-relational coordinate system is as follows:

[0122] ζ1=eta1-d1, ζ2=eta2-d2, ζ3=eta3-d1;

[0123] ζ4=eta4-d2, ζ5=eta5+d1, ζ6=eta6+d2;

[0124] where η i (i=1…6) is the corresponding state quantity in the hull coordinate system, d1 and d2 are coupled disturbance variables, which are formed by the roll, pitch and system state quantity coupling of the hull. The specific expression is:

[0125]

[0126] Step 2: construct an energy function to describe the energy change according to the total energy change rate of the marine crane system;

[0127] Among them, the total energy function of the crane system is:

[0128]

[0129] Taking the derivative, we get:

[0130]

[0131] in,

[0132] e=[ζ1 ζ2 ζ3 ζ4 e5 e6] T ;

[0133] e5 and e6 are the tracking errors of the cantilever heave angle and rotation angle, respectively, that is, e5 = ζ5-ζ 5r , e6=ζ6-ζ 6r ;

[0134] Step 3: construct a radial basis function neural network to estimate the lumped unknown dynamics in the system;

[0135] Among them, the unknown dynamic expression of the crane system is:

[0136]

[0137] In the formula, Υ 11 and 21 represents an unknown dynamic with an initial value of zero, and Υ 12 and 22 represents the unknown dynamics that are not initially zero, Δτ 51 and Δτ 61represents the items in Δτ5 and Δτ6 related to the system state quantity, while Δτ 52 and Δτ 62 Represents the items in Δτ5 and Δτ6 related to the mechanical structure of the system. The expression of the radial basis neural network is:

[0138]

[0139]

[0140] The update law expression of the neural network is:

[0141]

[0142] where Ψ1 and Ψ2 are positive definite diagonal matrices;

[0143] Step 4: Construct a nonlinear stable controller by combining radial basis neural network;

[0144] The expression of nonlinear stable controller is:

[0145]

[0146] Among them, δ π =2π / (ξ 6M -ξ 6m ) 2 , k p5 , k p6 , k d5 , k d6 , k s5 and k s6 All are control parameters and meet the following selection rules:

[0147]

[0148] Step 5: Use the designed controller to control and realize the positioning and anti-sway of the marine crane system.

[0149] Embodiment 2: Figure 1-Figure 4 As shown, a neural network-based nonlinear stability control system for offshore cranes includes a model building module, an energy analysis module, a radial basis neural network module, a controller design module and a nonlinear stability control module;

[0150] The model building module is configured to obtain the parameters of the ship crane system, consider the rolling and pitching motions of the ship hull, and establish a dynamic model of the ship crane system;

[0151] The energy analysis module is configured to construct an energy function to describe energy changes according to the total energy change rate of the ship crane system;

[0152] The radial basis neural network module is used to be configured to design an auxiliary function that can replace the position parameters of the control system;

[0153] The controller design module is configured to design a nonlinear stable controller according to the radial basis neural network and the system energy function analysis results;

[0154] The nonlinear stability control module is used to configure the control using a nonlinear stability controller to achieve positioning and sway elimination of the marine crane system.

[0155] Simulation test: The traditional PID controller and the control method of the present invention were tested using the Simulink tool in MATLAB. The two groups of traditional PID controllers were referred to as comparison group 1 and comparison group 2, and the two groups of nonlinear stability controllers using the present invention were referred to as example group 1 and example group 2. The test results are recorded in Table 1.

[0156] Table 1: Simulation comparison

[0157]

[0158] It can be seen from the above table that there is no obvious difference in the control performance of the method proposed in the present invention in different groups of system parameters. The cantilever can be fully positioned, and the hook load can converge to the equilibrium point after a cycle of swing, which shows that the method proposed in the present invention has good robustness.

[0159] The present invention uses the Lagrangian modeling method to establish a ship crane model, and constructs a Lyapunov energy function containing potential energy and kinetic energy based on the model. Subsequently, based on the first-order derivative of the energy function, the basic form of the adaptive controller is derived, and the designed radial basis function neural network is used to estimate the unknown dynamics of the system. The designed controller can realize the positioning of the cantilever and the load, and has a good suppression effect on the sway of the load.

[0160] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A nonlinear stability control method and system for offshore crane based on neural network, characterized in that: The following steps are involved: S1. Obtain the parameters of the ship crane system and establish a three-dimensional ship crane system dynamic model considering the roll and pitch motion of the ship; S2, constructing an energy function to describe the energy change according to the total energy change rate of the marine crane system; S3, constructing a radial basis neural network to estimate the aggregate unknown dynamics in the system; S4, construct a nonlinear stable controller by combining radial basis neural network; S5. Use the designed controller to control and realize the positioning and anti-sway of the marine crane system.

2. The neural network-based nonlinear stability control method for offshore cranes according to claim 1, characterized in that: In step S1, the Lagrangian modeling equation is used to construct a dynamic model of the marine crane system in combination with the wind resistance and wave disturbance existing in practical applications.

3. The nonlinear stability control method for offshore crane based on neural network according to claim 2 is characterized in that: The dynamic model of the marine crane system is expressed as: M=[m ij ] 6×6 ,C=[c ij ] 6×6 ,G=[g i ] 6×1 ; U=[0 0 0 0 τ5+Δτ5 τ6+Δτ6] T ; Where: M(ζ) is the mass matrix, is the centripetal-Coriolis force matrix, G(ζ) is the gravity vector, U is the input vector, F is the wind resistance and friction vector, ζ is the state variable, and are the velocity and acceleration of the state variables, τ5 and τ6 are the heave and rotation moments of the cantilever, Δτ5 and Δτ6 represent the nonlinear terms in the input dead zone, Represents the drag coefficient, f r5 and f r6 is the mechanical friction of the cantilever, f a ,δ5,f b and δ6 represents the static friction coefficient, κ a and κ b Represents the sliding friction coefficient.

4. The nonlinear stability control method for offshore crane based on neural network according to claim 3 is characterized in that: The internal elements of the matrix and vector have the following expressions: m 24 =m2l1l2(1+ζ2ζ4),m 25 =-(m1+m2)l1Lζ2S5; m 51 =m 15 ,m 52 =m 25 ,m 53 =m 35 ,m 55 =(m1+m2)L 2 +I y ; m 56 =-((m1+m2)l1Lζ2C5+m2l2Lζ4C5),m 54 =m 45 ; g4=m2l2gζ4, g2=(m1+m2)l1gζ2, g3=m2l2gζ3, g6=0; g5=-gS5(0.5M0L+(m1+m2)L-0.5M1L1); Where M0 and M1 are the masses of the ballast and boom, m0 and m1 are the masses of the hook and load, L, L1 and l1 are the lengths of the boom, ballast and rope, l2 is the distance between the hook and the load, and I x ,I y ,I z and I b are all parameters related to the moment of inertia, g is the acceleration due to gravity, C (*) and S (*) are the abbreviations of cos(*) and sin(*), τ5 and τ6 are the torques driving the boom to rise and fall and rotate, ζ1, ζ2, ζ3 and ζ4 are the swing angles of the hook and the load, ζ5 and ζ6 are the heave angle and rotation angle of the boom, respectively. i (i=1…6) are the corresponding angles in the inertial coordinate system. Their corresponding relationship with the angles in the crane non-relational coordinate system is as follows: ζ1=eta1-d1, ζ2=eta2-d2, ζ3=eta3-d1; ζ4=eta4-d2, ζ5=eta5+d1, ζ6=eta6+d2; where η i (i=1…6) is the corresponding state quantity in the hull coordinate system, d1 and d2 are coupled disturbance variables, which are formed by the coupling of the hull's roll, pitch and system state quantity. The specific expression is:

5. The neural network-based nonlinear stability control method for offshore cranes according to claim 1, characterized in that: In step S2, the total energy function of the crane system is: Taking the derivative, we get: in, e=[ζ1 ζ2 ζ3 ζ4 e5 e6] T ; e5 and e6 are the tracking errors of the cantilever heave angle and rotation angle, respectively, that is, e5 = ζ5-ζ 5r , e6=ζ6-ζ 6r .

6. The neural network-based nonlinear stability control method for offshore cranes according to claim 1, characterized in that: In step S3, the unknown dynamic expression of the crane system is: Among them, 11 and 21 represents an unknown dynamic with an initial value of zero, and Υ 12 and 22 represents the unknown dynamics that are not initially zero, Δτ 51 and Δτ 61 represents the items in Δτ5 and Δτ6 related to the system state quantity, while Δτ 52 and Δτ 62 Represents the items in Δτ5 and Δτ6 related to the mechanical structure of the system, and the expression of the radial basis neural network is:

7. The neural network-based nonlinear stability control method for offshore cranes according to claim 6, characterized in that: In step S3, the update law expression of the neural network is: where Ψ1 and Ψ2 are positive definite diagonal matrices.

8. The neural network-based nonlinear stability control method for offshore cranes according to claim 1, characterized in that: In step S4, the nonlinear stability controller expression is: in, k p5 , k p6 , k d5 , k d6 , k s5 and k s6 All are control parameters and meet the following selection rules: k s6 k p5 ,k p6 ,k d5 ,k d6 , 9. A nonlinear stability control system for offshore crane based on neural network, characterized in that: A neural network-based nonlinear stability control method for an offshore crane as described in any one of claims 1 to 8 is used, comprising a model building module, an energy analysis module, a radial basis neural network module, a controller design module and a nonlinear stability control module; The model building module is configured to obtain the parameters of the ship crane system, consider the roll and pitch motion of the ship hull, and establish a dynamic model of the ship crane system; The energy analysis module is configured to construct an energy function to describe energy changes according to the total energy change rate of the ship crane system; The radial basis neural network module is configured to design an auxiliary function that can replace the position parameters of the control system; The controller design module is configured to design a nonlinear stable controller according to the radial basis neural network and the system energy function analysis results; The nonlinear stability control module is configured to use a nonlinear stability controller to perform control to achieve positioning and sway elimination of the marine crane system.

Images