CN111745653A - A planning method for collaborative machining of hull outer plate surface forming based on dual manipulators - Google Patents

Abstract

The present invention proposes a planning method based on double manipulator arms for hull outer plate surface forming collaborative processing, which is used to solve the lack of a unified analysis and task planning model for the current multi-task planning method, and does not take the constraints into consideration of actual conditions. The problem in the task allocation process includes the following steps: S1, modeling analysis; S2, by determining the heating direction of the heating path, determining the start and end points of each heating path, and calculating the distance cost between the heating path and the heating path; S3 , selection and definition of variables; S4, determine the constraints; S5, determine the optimization goal of the double manipulator arm for the surface forming of the hull outer plate; S6, improve the artificial bee colony algorithm, and use the improved artificial bee colony algorithm to form the surface of the hull outer plate Allocation of machining tasks with dual robotic arms. The invention adopts the improved artificial bee colony algorithm to solve the task planning, and effectively solves the task assignment problem of the cooperative processing of the double mechanical arms.

Description

A planning method for collaborative machining of hull outer plate surface forming based on dual manipulators

technical field

The invention relates to the technical field of hull outer plate curved surface forming processing planning, in particular to a planning method for hull outer plate curved surface forming collaborative processing based on dual robotic arms.

Background technique

The hull is composed of complex and non-developable space curved surfaces, and the marine steel plate should be processed into the curved shape of the hull outer plate. At present, most of the methods adopted by shipyards in the world are the curved surface forming process of linear water-fired ship plates. The principle of this process is that the overall bending deformation of the steel plate is achieved by using the thermoelastic plastic deformation generated by the local steel plate being cooled by high temperature. Surface forming of hull outer plate is a process mode that uses the elastic-plastic deformation of the steel plate to complete the processing. Since the operation of the surface forming of the hull outer plate is dangerous and requires high experience of workers, this technology has always been developed by experience. Abundant skilled workers are done by hand. With the increasingly fierce competition in the modern shipbuilding market, more and more shipbuilding enterprises are considering how to improve the speed and efficiency of shipbuilding while ensuring quality. The biggest obstacle to speed and efficiency, so it is of great significance to realize the automation of the hull shell surface forming process.

In the field of hull outer plate surface forming industry, the manipulator has been used in shipbuilding due to its high working efficiency and reliable performance. However, in dealing with the complex and diverse tasks in the processing of the surface forming process of the hull outer plate, the single manipulator is increasingly showing insufficient capacity, while the double manipulator can complete complex work tasks through collaborative processing and improve work efficiency.

At present, the multi-task planning is mainly based on the classification of tasks at the workpiece level, and the motion of each manipulator in the system is calculated by the motion transformation of the relevant frames with a coordinated relationship, but there is still a lack of a unified analysis and task planning model. In the actual task assignment process, a large number of constraints have not been taken into account, such as the cooperation and competition of the dual manipulators in the actual processing of the hull outer plate and surface forming. Therefore, the solution is not accurate enough and cannot be applied well. into practice. Therefore, it is particularly important to develop a planning method for the co-processing of hull outer plate surface forming based on dual manipulators.

SUMMARY OF THE INVENTION

The present invention proposes a planning method for the co-processing of the hull outer plate surface forming based on double mechanical arms, so as to solve the lack of a unified analysis and task planning model for the current multi-task planning method, and does not take the constraints into account for the actual task assignment In the process, which leads to the problem of poor applicability of this solution, it is necessary to fully consider the constraints to effectively solve the task allocation problem of dual-manipulator collaborative processing, so that the task planning method can be well applied to practice.

The technical scheme of the present invention is realized as follows:

The planning method for the co-processing of the surface forming of the hull outer plate based on the dual manipulators specifically includes the following steps:

S1. Modeling analysis: According to the requirements of the co-processing process of the surface forming of the hull outer plate and the extraction of the heating path, various characteristics of the heating path are obtained;

S2, by determining the heating direction of the heating path, determining the start and end points of each heating path, and calculating the distance cost between the heating path and the heating path;

S3. Selection and definition of variables: including known constant definitions, variable definitions, and intermediate variable definitions of robotic arms;

S4, determine the constraints;

S5. Determine the optimization goal of the dual manipulators for the surface forming of the hull outer plate; according to various intermediate variables in the processing of the heating path of the dual manipulators, deduce the path cost of each manipulator during the working process, and obtain the idle time through the path cost. and processing time, and finally determine the time cost, and take the minimum total time as the optimization objective of the double manipulator for surface forming of the hull outer plate;

S6. The artificial bee colony algorithm is improved, and the improved artificial bee colony algorithm is used to allocate the processing tasks of the double manipulators for the surface forming of the hull outer plate.

To further optimize the technical solution, in the step S1, various features of the heating path include synchronous processing of dual robotic arms, competitive processing according to the optimization target, staggered processing time, and cooperative processing.

To further optimize the technical solution, the step S2 includes the following steps:

S21. If s ₁ and s ₂ are within the reachable range of one of the manipulators, they belong to the adjacent heating paths of the same manipulator. A ₁ and B ₁ are the endpoints of the heating path s ₁ , and A ₂ and B ₂ are heating paths. endpoint of path s2 _;

S22. Suppose that the heating path s ₁ is processed first, and then the heating path s ₂ is processed;

S23. If the machining direction of s ₁ is A ₁ →B ₁ , and the machining direction of s ₂ is ^A ₂ →B ₂ , then the machining order of the heating path is B ₁ →A ₁ ;

S24. If the machining direction of s ₁ is A ₁ →B ₁ , and the machining direction of s ₂ is B ₂ →A ₂ , the machining order of the heating path is B ₁ →B ₂ ;

S25. If the machining direction of s ₁ is B ₁ →A ₁ , and the machining direction of s ₂ is A ₂ →B ₂ , the machining order of the heating path is A ₁ →A ₂ ;

S26. If the machining direction of s ₁ is B ₁ →A ₁ , and the machining direction of s ₂ is B ₂ →A ₂ , the machining order of the heating path is A ₁ →B ₂ .

To further optimize the technical solution, in the step S3, the known variable definitions include the ID that defines the processing heating path, the length l of each processing heating path, the dual manipulators are defined as R ₁ and ^R ₂ , and the motion speed V of the manipulator is defined. ;

The variable definition includes defining the processing direction d of the heating path, the processing order x of the heating path, the robot arm ri to which the heating path belongs, the starting processing time t _i of the heating path _{, the finishing processing time Ti of the heating path, and the waiting time τ i of the} heating path;

The definition of the intermediate variables of the manipulator includes the heating path sequence Sj processed by each manipulator, the heating path length sequence L _j of each manipulator, the heating path processing direction sequence D _j of each manipulator, and the heating path processing of each manipulator. Start and end time σ _j , waiting time ψ _j of each robot arm.

Further optimizing the technical solution, in the step S4, the constraints include simultaneous processing constraints, safety time constraints, reachable space constraints, collision constraints, and kinematic constraints;

The synchronous machining constraint means that the start time and end time of the two heating paths that need to be synchronously processed are the same; the safety time constraint refers to the two heating paths that cannot be processed at the same time, and a period of time is required between processing; The space constraint means that each manipulator must be spatially accessible to all its assigned heating paths; the collision constraint refers to the collision between the manipulators and the collision between the manipulator and the processing material; the kinematics constraints are It means that the task planning must take into account the actual speed and acceleration constraints of the manipulator.

To further optimize the technical solution, in the step S5, the time cost includes the processing time, idling time and waiting time of the robotic arm.

To further optimize the technical solution, the step S6 includes the following steps:

S61. In the initial stage, randomly generate NP feasible solutions (x ₁ , x ₂ ,...,x _NP ) as the initial heating path for the movement of the processing robot arm. Each group of processing paths moved by the robot arm can be recorded as X _i , NP represents the number of food sources, and the number of leading bees and following bees is equal to the number of food sources;

S62. In the lead bee stage, each lead bee corresponds to a food source, and searches around it to obtain a new food source, and updates the processing path moved by the processing robotic arm in real time;

After the position of the food source is updated, the leading bee compares the nectar abundance of the candidate food source and the original food source. If the fitness value of the candidate food source is higher than that of the original food source, the candidate food source is used to replace the original food source, otherwise the original food source is maintained. The location of the food source does not change;

S63. Improvement of the artificial bee colony algorithm. The idea of the differential evolution search strategy will be introduced into the artificial bee colony algorithm to improve the search equation of the artificial bee colony algorithm;

S64. In the follower bee stage, according to the food source abundance information or fitness function value fed back by the lead bee, the follower bee selects the searched food source in a roulette manner, and searches the robotic arm according to the search equation in step S63 processing path;

S65. Execute the scout bee stage, that is, after all the lead bees and follower bees complete the search task, determine whether they are greater than the mining limit. If so, then the lead bees are transformed into scout bees, and randomly search for new food sources to replace the original food sources in the space. No, keep the original food source;

S66. After all the bees complete the search task, determine whether the termination condition is met. If the termination condition is met, record the processing path of the robotic arm, and the algorithm terminates. Otherwise, go to step S62, the bee colony starts to search again, and finally the bee colony is passed through the machine The heating paths in the working environment of the arm are connected in sequence, and the optimal solution for the dual manipulator task planning of hull outer plate surface forming is obtained.

To further optimize the technical solution, in the step S62, the search method is as follows:

v _ij =x _ij +r _ij (x _ij -x _kj )

Among them, i=1,2,...NP, v _ij is a candidate food source, x _kj is a randomly selected artificial bee, k∈(1,2,...,NP) and k≠i,j∈[1, 2,...,D] and is the dimension corresponding to the solution, all other variables will be inherited from the old food source, r _ij is a random number in [-1, 1], as the number of iterations increases, the adjacent The radius of the domain will gradually shrink, and finally the optimal solution will be obtained.

To further optimize the technical solution, in the step S62, comparing the nectar richness of the candidate food source and the original food source is to compare the fitness function value of each group of processing paths of the robotic arm, and calculate the fitness function of each food source by the following formula value:

To further optimize the technical solution, in the step S63, the search equation used is as follows:

v _ij =x _ij +η[(c ₁ x _pj +c ₂ x _qj +c ₃ x _rj -x _ij )+(x _bj -x _ij )]

Among them, x _pj , x _qj and x _rj are three randomly selected known solutions, c ₁ , c ₂ and c ₃ are determined by the fitness function values of the three known solutions, and c ₁ +c ₂ + c ₃ =1, x _bj is the current optimal food source location, T represents the current iteration number, and η is the differential variation factor as follows:

In the formula, a represents a constant, and the change degree of the differential variation factor η can be changed by adjusting the value of a.

Having adopted the above-mentioned technical scheme, the beneficial effects of the present invention are:

The invention models the multi-task planning of the dual manipulators as a complex multi-constraint TSP problem, and fully considers the cooperation and competition of the two manipulators in the actual processing of the hull hull plate surface forming. The characteristics of the plate surface forming heating path are analyzed, and the dual-manipulator collaborative processing task planning model is established.

The present invention firstly analyzes the process requirements of the hull outer plate surface forming process, carries out the dual manipulator task planning modeling analysis, and then optimizes the target with the shortest working time, and establishes the objective function of the hull outer plate surface forming dual manipulator coordinated processing task planning. Then, aiming at the problems of "premature", easy to fall into local optimum and slow convergence in the late iteration of the basic artificial bee colony algorithm, the differential evolution search strategy idea will be introduced into the artificial bee colony algorithm to improve the search equation of the artificial bee colony algorithm , which improves the original search equation into a complex polynomial form containing the current optimal solution. Since the current optimal solution guides the population, new candidate solutions are generated in the neighborhood of the optimal solution, which enhances the development ability of the algorithm.

In the early stage of the iteration under the influence of the differential evolution idea, the differential variation factor is relatively large, which is conducive to expanding the search space, improving the search ability of the algorithm, increasing the diversity of solutions, and improving the development ability of the population. In the later stage of the iteration, the differential variation factor gradually changes Small, it is beneficial for the algorithm to converge to the local optimal position, thereby improving the convergence accuracy. The improved artificial bee colony algorithm is used to solve the task planning, which effectively solves the task allocation problem of the collaborative processing of the dual manipulators, and fully considers the constraints, so that the task planning method can be well applied in practice. The automation of hull outer plate surface forming is of great significance.

Description of drawings

In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can also be obtained from these drawings without any creative effort.

Fig. 1 is the flow chart of the present invention;

FIG. 2 is a schematic diagram of the combination of the heating paths of the dual manipulators of the present invention;

FIG. 3 is a combination of adjacent heating paths for processing in the present invention;

Fig. 4 determines the adjacent heating paths of the processing direction for the present invention;

Fig. 5 is the schematic diagram of synchronous processing of the present invention;

FIG. 6 is a specific flow chart of the improved artificial bee colony algorithm of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

A planning method for co-processing of hull outer plate surface forming based on dual robotic arms, as shown in Figures 1 to 6, specifically includes the following steps:

S1, modeling analysis. According to the requirements of the co-processing technology of the surface forming of the hull outer plate and the extraction of the heating path, the characteristics of the heating path are obtained.

In step S1 , various features of the heating path include synchronous processing with dual manipulators, competitive processing according to the optimization target, staggered processing time, and cooperative processing.

Specifically, the characteristic relationships between the heating paths are as follows:

1) Some symmetrical heating paths require simultaneous processing of two robotic arms for the purpose of simultaneous deformation;

2) Some heating paths are in the common reachable area of the dual manipulators, and the manipulators will process them according to the optimization goal;

3) Some heating paths cannot be processed at the same time due to process requirements (such as preventing sheet deformation), so it is necessary to stagger each other's processing time;

4) Some heating paths are relatively long and exceed the motion range of any one robotic arm. At this time, two robotic arms are required to cooperate in processing;

5) Some heating paths are crossed, so it is necessary to stagger the processing time of the robotic arm.

As shown in Figure 2, step S1 specifically includes the following contents:

R ₁ and R ₂ represent the exclusive reachable space of the two manipulators respectively, I is the shared reachable space of the two manipulators, and the heating path in the shared reachable space is divided into competing heating paths according to the processing requirements , synchronous heating path, time exclusive heating path, cross heating path, the specific analysis is as follows:

S2. Determine the heating direction of the heating path. By determining the heating direction of each heating path, determining the start and end points of each heating path, and calculating the distance cost between the heating path and the heating path. In this step, after the machining direction of the heating path is determined, the distance cost w(s _i , s _j ) between the heating paths can be uniquely determined.

As shown in Figure 3 and Figure 4, step S2 includes the following steps:

S21. If s ₁ and s ₂ are within the reachable range of one of the manipulators, they belong to the adjacent heating paths of the same manipulator. A ₁ and B ₁ are the endpoints of the heating path s ₁ , and A ₂ and B ₂ are heating paths. endpoint of path _s2 .

S22. Suppose that the heating path s ₁ is processed first, and then the heating path s ₂ is processed. The heating path is shown in FIG. 2 .

S23. If the machining direction of s ₁ is A ₁ →B ₁ , and the machining direction of s ₂ is A ₂ →B ₂ , the machining sequence of the heating path is B ₁ →A ₁ .

S24. If the machining direction of s ₁ is A ₁ →B ₁ , and the machining direction of s ₂ is B ₂ →A ₂ , the machining sequence of the heating path is B ₁ →B ₂ .

S25. If the machining direction of s ₁ is B ₁ →A ₁ , and the machining direction of s ₂ is A ₂ →B ₂ , the machining sequence of the heating path is A ₁ →A ₂ .

S26. If the machining direction of s ₁ is B ₁ →A ₁ , and the machining direction of s ₂ is B ₂ →A ₂ , the machining order of the heating path is A ₁ →B ₂ .

In step S26, by determining the direction of each heating path, the start and end points of each heating path can be determined, and the distance cost between the heating path s ₁ and the heating path s ₂ can be calculated, which is denoted as w(s ₁ , s ₂ ) , after determining the processing direction of the heating path, the distance cost between the heating paths can be uniquely determined. As shown in FIG. 3 , when the machining direction of the heating path of s ₁ is determined as A ₁ →B ₁ , and the machining direction of the heating path of s ₂ is A ₂ →B ₂ , the two heating paths s ₁ and s ₂ can be determined. The machining sequence of the path is B ₁ →A ₁ .

S3. Selection and definition of variables: including known constant definitions, variable definitions, and intermediate variable definitions of robotic arms.

In step S3, the known variable definition includes defining the ID of the processing heating path, the length l of each processing heating path, defining the dual robotic arms as R ₁ and R ₂ , and defining the moving speed V of the robotic arm.

The variable definition includes defining the heating path processing direction d, the heating path processing order x, the robot arm ri to which the heating path belongs, the heating path start processing time _ti , the heating path ending processing time Ti _, and the heating path waiting time τ _{i .}

The definition of the intermediate variables of the manipulator, the heating path sequence processed by each manipulator, the length sequence of the heating path of each manipulator, the processing direction sequence of the heating path of each manipulator, the processing start and end time of the heating path of each manipulator, Waiting time for each robotic arm.

Step S3 specifically includes the following steps:

S31. Definition of known constants. The main constants are defined in the following table:

S32, variable definition. The main variables are defined in the following table:

S33, the definition of the intermediate variable of the manipulator. Intermediate variables are defined in the following table:

S4, determine the constraints. In step S4, the constraints include simultaneous processing constraints, safe time constraints, reachable space constraints, collision constraints, and kinematic constraints.

The synchronous machining constraint means that the start time and end time of the two heating paths that need to be synchronously processed are the same; the safety time constraint refers to the two heating paths that cannot be processed at the same time, and a period of time is required between processing; The space constraint means that each manipulator must be spatially accessible to all its assigned heating paths; the collision constraint refers to the collision between the manipulators and the collision between the manipulator and the processing material; the kinematics constraints are It means that the task planning must take into account the actual speed and acceleration constraints of the manipulator.

The constraints included in the objective optimization in step S4 are as follows:

Step S4 includes the following steps:

的确定。为了保证加工工艺得要求，有些加热路径需要同步加工，如图4所示为同步加工示意图。同步加工要求两条加热路径的开始时间、结束时间一致，对于同步加热路径s_i,s_j，要求t_i＝t_j,T_i＝T_j，其中i,j∈S。S41. Simultaneous processing constraints

ok. In order to ensure the requirements of the processing technology, some heating paths need to be processed synchronously, as shown in Figure 4 for the schematic diagram of synchronous processing. Simultaneous processing requires the same start time and end time of the two heating paths. For the synchronous heating paths s _i , s _j , it is required that t _i =t _j , Ti =T _j , where _i ,j∈S.

的确定。由于船体外板曲面成形加工工艺得要求，防止局部温度过高，导致板材的材质受到损坏，某些加热路径不能在在同一时间内进行加工，加热路径的加工时中间需要间隔一段时间。所以在出现这些加热路径同时加工时，则需要其中一台机械臂必须停下来等待其中一台机械臂加工完成后在进行加工。在实际的加工中，对于时间互斥的加热路径，需要满足式(1)。S42. Safety time-constrained

ok. Due to the requirements of the surface forming process of the hull outer plate, the local temperature is too high to prevent the material of the plate from being damaged. Some heating paths cannot be processed at the same time, and the processing of the heating paths needs to be separated for a period of time. Therefore, when these heating paths are processed at the same time, one of the robotic arms must stop and wait for one of the robotic arms to complete the processing before processing. In actual processing, it is necessary to satisfy Equation (1) for the mutually exclusive heating paths in time.

T _i <t _j ‖T _j <t _i (i,j∈S) Equation (1)

的确定。在双机械臂的任务规划中每一台机械臂，对于它分配所有加热路径必须是空间可达的，对于衡量一条加热路径是否可达，采用的方法是在对加热路径信息提取时所获得的一系列离散的加热路径特征点，然后根据机械臂的D-H参数和逆解，计算每一个点是否在加热路径所分配机械臂的可达空间内，如果每一点都在，此加热路径就在所属机械臂的可达空间内。对于合作加热路径，可以将其分为两段加热路径，然后分别在对应机械臂可达空间即可。记加热路径s_i所属机械臂可达空间为R_j(q)，q为机械臂的关节的运动范围，需要满足：s_i∈R_j(q)，其中i∈S，j＝1、2。S43. Reachable space constraints

ok. In the task planning of the dual manipulators, each manipulator must be spatially accessible for all the heating paths allocated to it. To measure whether a heating path is reachable, the method used is obtained when extracting the heating path information. A series of discrete heating path feature points, and then according to the DH parameters of the robot arm and the inverse solution, calculate whether each point is within the reachable space of the robot arm assigned by the heating path. If each point is there, the heating path belongs to within the reach of the robotic arm. For the cooperative heating path, it can be divided into two sections of heating path, and then the corresponding robot arm can reach the space respectively. Denote the reachable space of the robotic arm to which the heating path _si belongs as R _j (q), and q is the motion range of the joint of the robotic arm, which needs to satisfy: s _i ∈ R _j (q), where i ∈ S, j=1, 2 .

的确定。在实际的加工中机械臂和代加工材料都是刚性物体，一旦发生碰撞将造成机械臂损坏等严重后果，船体外板曲面成形双机械臂的碰撞包括机械臂之间的碰撞和机械臂和加工材料之间碰撞。令加工材料所占用空间为R_wp，机械臂1此时所在的空间为R₁(q₁)，q₁为机械臂1当前关节向量，机械臂2此时所在的空间为R₂(q₂)，q₂为机械臂2当前关节向量，需要满足式(2)。S44. Collision constraint

ok. In actual processing, both the manipulator and the processing material are rigid objects. Once a collision occurs, it will cause serious consequences such as damage to the manipulator. The collision of the double manipulators for the surface forming of the hull outer plate includes the collision between the manipulators and the manipulator and the processing. collisions between materials. Let the space occupied by the processing material be R _wp , the space where the robot arm 1 is at this time is R ₁ (q ₁ ), q ₁ is the current joint vector of the robot arm 1, and the space where the robot arm 2 is at this time is R ₂ (q ₂ ), q ₂ is the current joint vector of robot arm 2, which needs to satisfy formula (2).

R _wp ∩R ₁ (q ₁ )=φ, R _wp ∩R ₂ (q ₂ )=φ, R ₁ (q ₁ )∩R ₂ (q ₂ )=φ Equation (2)

的确定，令v_ij(t)为机械臂i第j关节当前速度，V_ij为机械臂i第j关节速度范围，a_ij(t)为机械臂i第j关节当前加速度，A_ij为机械臂i第j关节加速度范围，则需要满足式(3)，式中i＝1、2，j∈机械臂i关节角个数。S45, kinematic constraints

to determine, let v _ij (t) be the current speed of the jth joint of the manipulator i, V _ij be the speed range of the jth joint of the manipulator i, a _ij (t) be the current acceleration of the jth joint of the manipulator i, and A _ij be the mechanical The acceleration range of the jth joint of arm i needs to satisfy formula (3), where i=1, 2, j∈ the number of joint angles of arm i.

v _ij (t)∈V _ij , a _ij (t)∈A _ij Formula (3).

S5. Determine the optimization goal of the dual manipulators for the surface forming of the hull outer plate; according to various intermediate variables in the processing of the heating path of the dual manipulators, deduce the path cost of each manipulator during the working process, and obtain the idle time through the path cost. and processing time, and finally determine the time cost, taking the minimum total time as the optimization objective of the double manipulator for surface forming of the hull outer plate. In this step, first determine the path cost U _Wj and time cost U _Tj of each manipulator in the working process, and then determine the optimization objective U of the dual manipulator task planning.

In step S5, the time cost includes the processing time, idle time and waiting time of the robot arm.

Step S5 includes the following steps:

S51. According to various intermediate variables during the processing of the heating path of the dual manipulators, the path cost U _Wj and the time cost U _Tj of each manipulator in the working process can be deduced, as shown in formula (4) and formula (5), where The distance cost is the sum of the path traveled by the robot arm. After the path of the robot arm is determined, the processing time and the idle time can be determined according to the processing speed and idle speed of the robot arm.

S52. The optimization goal of the task planning of the dual manipulators is determined. For a certain heating path combination, the processing time is determined by the manipulator with the longest processing time among the two manipulators, and the processing time of the manipulator is the entire processing time. The time is as in formula (6), and the optimization goal is to minimize U, that is, to find minU.

U=max _{j=1, 2} {U _Tj (r,x,d,l,v,τ)} Equation (6)

S6. The basic artificial bee colony algorithm has problems such as "premature", easy to fall into local optimum and slow convergence in the later stage of iteration. Therefore, the artificial bee colony algorithm is improved, and the improved artificial bee colony algorithm is used to form the hull outer plate surface Allocation of machining tasks for robotic arms. The improved artificial bee colony algorithm is used to solve the task planning. The algorithm flowchart is shown in Figure 5. The idea of difference is introduced in the follower bee stage of the artificial bee colony algorithm, and the search equation of the artificial bee colony algorithm is improved. The original search equation It is improved into a complex polynomial form containing the current optimal solution. By improving the search method in the follower bee stage, the multi-task planning of the hull hull plate surface forming dual manipulator is optimized, and the optimal dual manipulator processing task assignment method is obtained.

Step S6 includes the following steps:

S61. In the initial stage, randomly generate NP feasible solutions (x ₁ , x ₂ ,...,x _NP ) as the initial heating path for the movement of the processing robot arm. Each group of processing paths moved by the robot arm can be recorded as X _i , As in formula (5), where NP represents the number of food sources, and the number of leading bees and follower bees is equal to the number of food sources. Each element in each set of moving machining paths X _i corresponds to the heating path in the established task planning model.

X _i =(x _i1 ,x _i2 ,...,x _iD )(i=1,2,...,NP) Formula (5).

S62. In the lead bee stage, each lead bee corresponds to a food source, and searches around it to obtain a new food source, and updates the processing path moved by the processing robotic arm in real time. The search method is as shown in formula (6).

v _ij =x _ij +r _ij (x _ij -x _kj ) Equation (6)

where i=1,2,...NP, v _ij is a candidate food source, x _kj is a randomly selected artificial bee, k∈(1,2,...,NP) and k≠i,j∈[1,2 ,…,D] and is the dimension corresponding to the solution, all other variables will be inherited from the old food source, r _ij is a random number in [-1, 1], with the increasing number of iterations, the neighborhood The radius will gradually shrink, and finally the optimal solution will be obtained.

After the position of the food source is updated, the leading bee compares the nectar abundance of the candidate food source and the original food source. If the fitness value of the candidate food source is higher than that of the original food source, the candidate food source is used to replace the original food source, otherwise the original food source is maintained. The location of the food source does not change.

In step S62, comparing the nectar richness of the candidate food source and the original food source is to compare the fitness function value of each group of processing paths of the robotic arm, and the fitness function value of each food source is calculated by the following formula:

S63. Improvement of the artificial bee colony algorithm. In the original artificial bee colony algorithm, the following bees and the leading bees both use the same search strategy. The search strategy has better global search ability, but ignores the local search performance of the algorithm. In order to further Improve the search ability of the artificial bee colony algorithm, increase the diversity of the population, introduce the idea of mutation in the differential evolution algorithm into the search strategy, and obtain a new search equation such as formula (8);

v _ij =x _ij +η[(c ₁ x _pj +c ₂ x _qj +c ₃ x _rj -x _ij )+(x _bj -x _ij )] Equation (8)

where x _pj , x _qj and x _rj are three randomly selected known solutions, c ₁ , c ₂ and c ₃ are determined by the fitness function values of the three known solutions, and c ₁ +c ₂ +c ₃ = 1, x _bj is the current optimal food source position, T represents the number of current iterations, and η is the differential variation factor as in formula (9).

In formula (9), a represents a constant, and the variation degree of the differential variation factor η can be changed by adjusting the value of a.

It can be seen from the new search equation that determining the values of c ₁ , c ₂ and c ₃ according to the fitness function values of the three _randomly selected known solutions greatly increases the diversity of the population; Guide the population and generate new candidate solutions in the neighborhood of the optimal solution, which can improve the development ability of the algorithm. The differential variation factor η gradually decreases with the increase of the number of iterations. The smaller the differential variation factor is, the smaller the search range of the bee colony is. In the early stage of iteration, the differential variation factor is relatively large, which is conducive to expanding the search space, improving the global search ability of the algorithm, increasing the diversity of solutions, and improving the development ability of the algorithm. Convergence to the local optimal position, thereby improving the convergence accuracy.

S64. In the follower bee stage, the follower bee selects the searched food source in a roulette manner according to the food source abundance information or the fitness function value fed back by the lead bee, and searches the robotic arm according to the search equation in step S63 processing path.

S65. Execute the reconnaissance bee stage, that is, after all the lead bees and follower bees complete the search task, determine whether it is greater than the mining limit limit. If so, the lead bee will be transformed into a scout bee, and a new food source will be randomly searched in space to replace the original food source. If not, the original food source will be retained.

S66: After all the bees complete the search task, determine whether the termination condition is reached. The termination condition is reaching the maximum number of iterations or reaching the target accuracy. If the termination condition is met, the processing path of the robotic arm is recorded, and the algorithm is terminated; otherwise, go to step S62, the bee colony starts to search again, and finally the bee colony is connected in sequence through the heating paths in the working environment of the robotic arm. It is the optimal solution for the task planning of dual manipulators for hull outer plate surface forming.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of the present invention. within the scope of protection.

Claims (10)

1. The planning method for the forming and collaborative processing of the hull plate curved surface based on the double mechanical arms is characterized by comprising the following steps:

s1, modeling analysis: according to the requirement of the ship hull plate curved surface forming co-processing technology and the extraction of the heating path, obtaining various characteristics of the heating path;

s2, determining the starting and ending points of each heating path by determining the heating direction of the heating path, and calculating the distance cost between the heating path and the heating path;

s3, selection and definition of variables: the method comprises the steps of known constant definition, variable definition and mechanical arm intermediate variable definition;

s4, determining constraint conditions;

s5, determining a double-mechanical-arm optimization target for forming the curved surface of the hull plate; deducing the path cost of each mechanical arm in the working process according to various intermediate variables in the processing process of the heating paths of the two mechanical arms, calculating the idle running time and the processing time according to the path cost, finally determining the time cost, and forming the optimization target of the two mechanical arms for the hull outer plate curved surface by using the minimum total time;

s6, improving the artificial bee colony algorithm, and distributing the processing tasks of the double mechanical arms for forming the curved surface of the hull plate by adopting the improved artificial bee colony algorithm.

2. The planning method for forming and co-processing the curved surface of the hull plate based on the two robot arms as claimed in claim 1, wherein the characteristics of the heating path in step S1 include synchronous processing of the two robot arms, competing processing according to an optimization target, staggering processing time, and co-processing.

3. The planning method for the curved surface forming and collaborative processing of the ship hull plate based on the double mechanical arms as claimed in claim 1, wherein the step S2 includes the steps of:

s21, if S₁And s₂Is within reach of one of the robots, belonging to adjacent heating paths of the same robot, A₁、B₁Is a heating path s₁End point of (A)₂、B₂Is a heating path s₂The endpoint of (1);

s22, assuming that the heating path S is processed first₁Reprocessing s₂；

S23, if S₁The machine direction of (A)₁→B₁，s₂The machine direction of (A)₂→B₂The processing order of the heating paths is B₁→A₁；

S24, if S₁The machine direction of (A)₁→B₁，s₂In the machine direction of B₂→A₂The processing order of the heating paths is B₁→B₂；

S25, if S₁In the machine direction of B₁→A₁，s₂The machine direction of (A)₂→B₂The processing order of the heating paths is A₁→A₂；

S26, if S₁In the machine direction of B₁→A₁，s₂In the machine direction of B₂→A₂The processing order of the heating paths is A₁→B₂。

4. The planning method for curved surface forming and collaborative processing of ship hull plates based on two robots as claimed in claim 1, wherein in the step S3, the known variable definitions include ID for defining processing heating paths, length l for each processing heating path, and R for defining two robots₁And R₂Defining the movement speed V of the mechanical arm;

the variable definition comprises the definition of the heating path processing direction d, the heating path processing sequence x and the mechanical arm r to which the heating path belongs_iHeating path start processing time t_iHeating path end processing time T_iHeating path waiting time τ_i；

The definition of the middle variable of the mechanical arm comprises a heating path sequence Sj processed by each mechanical arm and a heating path length sequence L of each mechanical arm_jHeating path processing direction sequence D of each mechanical arm_jHeating path processing start and end time sigma of each mechanical arm_jWaiting time psi for each robot arm_j。

5. The planning method for the curved surface forming and collaborative processing of the ship hull plate based on the double mechanical arms as recited in claim 1, wherein in the step S4, the constraint conditions include a synchronous processing constraint condition, a safe time constraint condition, a reachable space constraint condition, a collision constraint condition, and a kinematic constraint condition;

the synchronous processing constraint condition means that the start time and the end time of two heating paths needing synchronous processing are consistent; the safe time constraint condition refers to that two heating paths which cannot be processed at the same time need to be separated by a period of time in the middle of processing; the reachable space constraint means that each robot must be space-reachable for all its assigned heating paths; the collision constraint condition refers to the collision between the mechanical arms and the processing material; the kinematics constraint condition means that the actual speed and acceleration constraints of the mechanical arm must be considered in task planning.

6. The planning method for the curved surface forming and collaborative processing of the ship hull plate based on the double mechanical arms as claimed in claim 1, wherein the time cost in step S5 includes processing time, idle running time and waiting time of the mechanical arm.

7. The planning method for the curved surface forming and collaborative processing of the ship hull plate based on the double mechanical arms as claimed in claim 1, wherein the step S6 includes the steps of:

s61, in the initial stage, NP feasible solutions (x) are randomly generated₁,x₂,...,x_NP) As the initial heating path for the movement of the processing robot, each set of processing paths for the movement of the robot may be denoted as X_iNP represents the number of food sources, and the number of leading bees and following bees is equal to the number of food sources;

s62, leading bees, wherein each leading bee corresponds to one food source, a new food source is obtained by searching around the leading bee according to the formula, and the processing path moved by the processing mechanical arm is updated in real time;

after the positions of the leading bees in the food sources are updated, comparing the nectar richness degree of the candidate food sources with that of the original food sources, if the fitness value of the candidate food sources is higher than that of the original food sources, replacing the original food sources with the candidate food sources, otherwise, maintaining the positions of the original food sources unchanged;

s63, improving the artificial bee colony algorithm, namely, improving a search equation of the artificial bee colony algorithm by introducing a differential evolution search strategy idea into the artificial bee colony algorithm;

s64, a bee following stage, namely selecting the searched food source in a roulette mode by the following bee according to the information of the richness degree of the food source fed back by the leading bee or the fitness function value, and searching the processing path of the mechanical arm according to the search equation in the step S63;

s65, executing a scout bee stage, namely judging whether all leading bees and following bees finish searching tasks, if so, converting the leading bees into scout bees, randomly searching a new food source in the space to replace the original food source, and if not, reserving the original food source;

s66, after all bees finish the search task, judging whether a termination condition is reached, if the termination condition is met, recording the processing path of the mechanical arm, terminating the algorithm, otherwise, turning to the step S62, starting to search again for the bee colony, and finally connecting the bee colony sequentially through all heating paths in the mechanical arm working environment to obtain the optimal solution for the task planning of the double mechanical arms for forming the hull outer plate curved surface.

8. The planning method for the curved surface forming and cooperative processing of the ship hull plate based on the double mechanical arms as claimed in claim 7, wherein in the step S62, the searching manner is as follows:

v_ij＝x_ij+r_ij(x_ij-x_kj)

wherein i ═ 1, 2.. NP, v_ijIs a candidate food source, x_kjIs a randomly selected artificial bee, k ∈ (1,2,.., NP) and k ≠ i, j ∈ [1,2,..., D)]And is the dimension of the solution, all other variables will be inherited from the old food source, r_ijIs [ -1,1 [ ]]The radius of the neighborhood is gradually reduced along with the continuous increase of the iteration number, and the optimal solution is finally obtained.

9. The planning method for collaborative processing of curved surface forming of ship hull plate based on two robots as claimed in claim 7, wherein in step S62, the degree of richness of nectar of the candidate food sources and the original food sources is compared with the fitness function value of each group of processing paths of the comparison robot, and the fitness function value of each food source is calculated by the following formula:

10. the planning method for the curved surface forming and cooperative processing of the ship hull plate based on the double mechanical arms as claimed in claim 7, wherein the search equation adopted in the step S63 is as follows:

v_ij＝x_ij+η[(c₁x_pj+c₂x_qj+c₃x_rj-x_ij)+(x_bj-x_ij)]

wherein x is_pj、x_qjAnd x_rjIs three known solutions chosen at random，c₁、c₂And c₃Is determined by the fitness function value of three known solutions, and c₁+c₂+c₃＝1，x_bjIs the current optimal food source position, T represents the number of current iterations, η is the differential variation factor as follows:

in the formula, a represents a constant, and the degree of variation of the differential variation factor η can be changed by adjusting the value of a.

Images