CN104239621A - Line heating automatic forming method based on curve surface unfolding - Google Patents

Abstract

The invention discloses a line heating automatic forming method based on curve surface unfolding. The method comprises the following steps that on the basis of the existing boat plate Tribon data, a curve surface unfolding method is adopted for calculating the path arrangement of a heating line; then, proper processing technical parameters are selected from a line heating expert system for forming processing. After the primary processing forming, the formed curve surface is subjected to measurement and fitting by using a computer visual technology and is inputted into a processing system, then, a boat plate is unfolded, the unfolding surface is compared with the unfolding surface required in the design drawing in shrinkage amount, and if the comparison result does not reach the forming quantity requirement and is beyond the error range, the processing system carries out parameter selection again; if the boat plate unfolding quantity comparison error is within the required range, the forming reaches the requirement. The method provided by the invention has the advantages that the line heating automatic forming is realized, and the forming rate and the production efficiency of the boat plate are improved.

Description

Based on the flame forming plate automatic forming method of Doubly curved surfaces

Technical field

The present invention relates to a kind of flame forming plate forming method of Ship body, particularly relate to a kind of flame forming plate automatic forming method based on Doubly curved surfaces, belong to shipbuilding technical field.

Background technology

Ship body is welded by the not extensible space curved surface plate of a lot of block complexity, flame forming plate (Line Heating) forming technology of Double curve degree ship hull plate originates from Japan in twentieth century five sixties, because its processing is fast, flexible operation, does not need other equipment and be suitable for the advantage such as formed machining of complicated shape, extensively be applied in shipyard all over the world rapidly, and having become the main processing mode of Double curve degree outside plate, Double curve degree ship hull plate nearly all is at present all adopt flame forming plate process forming.But for a long time, this technique is always by workman's manual operations by rule of thumb.Artificial experience operate not unit consumption time, effort, and forming quality is poor, affects ship building period.Along with the development of Modern Shipbuilding technology and the transformation of Shipbuilding Mode, the development of flame forming plate automatic processing device has become in face of people one of pendulum task anxious to be resolved.

Nearly tens, the scientific research institution of lot of domestic and international and colleges and universities are in the research being devoted to flame forming plate automation equipment, but never ripe automation equipment is applied in actual shipyard, the automation equipment especially realizing complete line heating shaping is few especially.Technical bottleneck main is at present that the research of the research of heat distortion mechanism and heating parameters is not mature enough, and the detection mode of the deck of boat has also become the obstacle of serious obstruction bent plate automation equipment exploitation after processing, be not suitable for the closed-loop control of robotization processing, need to change yet.Therefore, research flame forming plate automatic forming method, the closed loop processing realizing complete line hot briquetting controls, and has extraordinary application and promotional value.

Summary of the invention

The object of the present invention is to provide a kind of flame forming plate automatic forming method based on Doubly curved surfaces, call at the molding mode solving the pure manual processing of current shipbuilding industry flame forming plate technique, realize the robotization processing of flame forming plate, improve the detection mode after deck of boat processing, save the complex detection instrument that conventional shape detects, realize complete line heating automatic closed loop controlled working; Improve the condition of work of workman, reduce labor intensity, improve the shaping rate of the deck of boat and the production efficiency of enterprise.

Object of the present invention is achieved by the following technical programs:

Based on a flame forming plate automatic forming method for Doubly curved surfaces, comprise the following steps:

1) importing of real deck of boat Tribon data;

2) the method Developing Surface of Doubly curved surfaces is applied;

3) the crack place of plate launched is carried out to the calculating of amount of contraction, utilize inference machine computing system to calculate and provide heater wire paths arrangement;

4) according to the heater wire arranged, system of processing selects the machined parameters that can reach designing requirement inside flame forming plate expert system, determine initial heating position by the heater wire path provided, drive heating head and water-cooling apparatus to realize the robotization processing of flame forming plate;

5) first after completion of processing, treat that steel plate is cooled to normal temperature, by camera filming apparatus and image processing system module, matching is carried out to the curve form of the plate after processing;

6) on the basis of matching, again One step development is carried out to the plate after processing, after expansion, calculates amount of contraction, and with initial deployment amount of contraction carries out difference comparsion, if amount of contraction does not reach designing requirement, and outside permissible error scope, then forward step 3 to); If the error between them within the range of permission, terminate processing.

Object of the present invention can also be realized further by following technical measures:

The aforementioned flame forming plate automatic forming method based on Doubly curved surfaces, wherein step 4) described flame forming plate expert system summed up by a large amount of production experience knowledge, the technological parameter of the processing of the common deck of boat integrated and optimizes; The reliability model utilizing finite element emulation software to set up processes the emulated data that flame forming plate forming process is carried out and optimizes; One that a sets up expert system knowledge base in conjunction with the deflection of working process parameter, deck of boat oneself factor, environmental factor and most soleplate.

The aforementioned flame forming plate automatic forming method based on Doubly curved surfaces, wherein step 2) curved surface unwinding method be:

The expression carrying out uniform B-Spline to the ship hull plate imported carrys out the curved surface of this real deck of boat of matching, minimum principle is changed based on keeping the distance launched on the curved surface of front and back between all net points, initial delta expansion is carried out to curved surface, using initial expanding data as iteration initial value, be the data that overall artificial fish-swarm algorithm carries out that iterative obtains all breaking up points with the multirow improved.

The aforementioned flame forming plate automatic forming method based on Doubly curved surfaces, wherein curved surface unwinding method is:

That after keeping the distance in adjacent mesh in the pre-deployed, variable quantity is minimum as far as possible based on extreme value thinking method of deploying cardinal rule, point M (x, y, z) be the point on original curved surface, its data can calculate through fitting, point m (x, y) for launching the rear number of planes strong point corresponding with it, the expansion of curved surface is according to raw data points M (x, y, z) calculate the process of m (x, y), then approximate expansion problem can be grouped into following extreme-value problem:

$s=\frac{1}{2}\underset{\mathrm{ij}}{\mathrm{\Σ}}\underset{N}{\mathrm{\Σ}}{(L_{\mathrm{ij}}-l_{\mathrm{ij}})}^2=\min ---\left(1\right)$

In formula: L _ij=| M _im _j| be the distance of its original curved surface any two points;

L _ij=| m _im _j| for launching the distance of latter corresponding 2;

N is adjacent with M (x, y, z) coordinate a little;

To a M (x, y, z), N is the coordinate of eight points around it, the distance L of some M (x, y, z) and adjacent eight points before namely launching _ij, with the distance l putting m (x, y) and adjacent eight points after expansion _ij, variable quantity is minimum;

This equation is about variable x _ijand y _ijnonlinear equation, solve the coordinate that this equation can obtain launching rear each point.Can obtain equation analysis, ask the necessary condition of s minimum value to be that s is to x _ijand y _ijpartial derivative be 0, that is:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=0\end{array}\right.---\left(2\right)$

Respectively to x _ijand y _ijdifferentiate, above formula becomes:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](x_{\mathrm{ij}}-x_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](y_{\mathrm{ij}}-y_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\end{array}\right.---\left(3\right)$

In formula, ab is the coordinate of some M (x, y, z) consecutive point.

The aforementioned flame forming plate automatic forming method based on Doubly curved surfaces, the multirow wherein improved is that overall artificial fish-swarm algorithm step is as follows:

(1) initialization operation is carried out to parameter, initialization Artificial Fish population number, artificial and initial position, Artificial Fish visual field Visual, Artificial Fish step-length Step, maximum iteration time, maximum attempts Try_number, crowding factor parameter;

(2) fitness of every bar Artificial Fish is calculated, and will there is the Artificial Fish state recording of optimal objective function to bulletin board;

(3) behavior that every bar Artificial Fish will carry out is evaluated, select the most suitable behavior campaign of Artificial Fish instantly;

(4) behavior selected by Artificial Fish execution, and upgrade the status information of Artificial Fish;

(5) state with the Artificial Fish of global optimum is upgraded;

(6) algorithm cycling condition is evaluated, if meet termination condition, then export optimal value, terminate program and run, otherwise jump to step (2) and again circulate, until algorithm termination condition is satisfied.

Compared with prior art, the invention has the beneficial effects as follows: the present invention solves that the pure manual processing mode of current shipbuilding industry flame forming plate technique is consuming time, effort, and forming quality is poor, affect the problem of ship building period, realize the robotization processing of flame forming plate, improve the detection mode after deck of boat processing, save the complex detection instrument that conventional shape detects, realize complete line heating automatic closed loop controlled working; Improve the condition of work of workman, reduce labor intensity, improve the shaping rate of the deck of boat and the production efficiency of enterprise.

Accompanying drawing explanation

Fig. 1 is flame forming plate automatic forming method FB(flow block) of the present invention;

Fig. 2 is schematic diagram before and after Doubly curved surfaces, and Fig. 2 (A) is original curved surface, and Fig. 2 (B) is Developing Surface;

Fig. 3 is the multirow improved is overall artificial fish-swarm algorithm process flow diagram.

Embodiment

Below in conjunction with the drawings and specific embodiments, the invention will be further described.

As shown in Figure 1, the technological process that the present invention is based on the flame forming plate automatic forming system that curved surface is expressed is: 1) import certain real deck of boat Tribon data; 2) expression carrying out uniform B-Spline to the ship hull plate imported carrys out the curved surface of this real deck of boat of matching; 3) carrying out initial delta expansion to curved surface, using initial expanding data as iteration initial value, is the data that overall fish-swarm algorithm carries out that iterative can obtain all breaking up points with the multirow improved; 4) the crack place of plate launched is carried out to the calculating of amount of contraction, utilize inference machine computing system to calculate and provide heater wire paths arrangement; 5) according to the heater wire arranged, system of processing selects the machined parameters that can reach designing requirement inside flame forming plate expert system, determine initial heating position by the heater wire path provided, drive heating head and water-cooling apparatus to realize the robotization processing of flame forming plate; 6) first after completion of processing, treat that steel plate is cooled to normal temperature, by camera filming apparatus and image processing system module, matching is carried out to the curve form of the plate after processing; 7) on the basis of matching, again One step development is carried out to the plate after processing, after expansion, calculates amount of contraction, and with initial deployment amount of contraction carries out difference comparsion, if amount of contraction does not reach designing requirement, and outside permissible error scope, then forward step 4 to); If the error between them within the range of permission, terminate processing.

Here the expansion principle that take is that the distance change keeping launching on the curved surface of front and back between all net points is minimum, and it is zero that this criterion is equivalent to remaining nodal force vector in Finite Element Method.Compared with the direct method of triangle expansion etc., this method asking extreme-value problem, is essentially to ask and launches the minimum optimization method of rear error.

That after keeping the distance in adjacent mesh in the pre-deployed, variable quantity is minimum as far as possible based on extreme value thinking method of deploying cardinal rule.As shown in Figure 2, some M (x, y, z) is the point on original curved surface, and its data can calculate through fitting.Point m (x, y) is number of planes strong point corresponding with it after launching.The expansion of curved surface is the process calculating m (x, y) according to raw data points M (x, y, z), then approximate expansion problem can be grouped into following extreme-value problem:

$s=\frac{1}{2}\underset{\mathrm{ij}}{\mathrm{\Σ}}\underset{N}{\mathrm{\Σ}}{(L_{\mathrm{ij}}-l_{\mathrm{ij}})}^2=\min ---\left(1\right)$

In formula: L _ij=| M _im _j| be the distance of its original curved surface any two points

L _ij=| m _im _j| for launching the distance of latter corresponding 2

N is adjacent with M (x, y, z) coordinate a little

To putting M (x, y, z) in Fig. 2, N is the coordinate of eight points around it, the distance L of some M (x, y, z) and adjacent eight points before namely launching _ij, with the distance l putting m (x, y) and adjacent eight points after expansion _ij, variable quantity is minimum.

This equation is about variable x _ijand y _ijnonlinear equation, solve the coordinate that this equation can obtain launching rear each point.Can obtain equation analysis, ask the necessary condition of s minimum value to be that s is to x _ijand y _ijpartial derivative be 0, that is:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=0\end{array}\right.---\left(2\right)$

Respectively to x _ijand y _ijdifferentiate, above formula becomes:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](x_{\mathrm{ij}}-x_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](y_{\mathrm{ij}}-y_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\end{array}\right.---\left(3\right)$

In formula, ab is the coordinate of some M (x, y, z) consecutive point.

Solving for equation, adopt the multirow improved to be overall artificial fish-swarm algorithm, this algorithm all has significantly improved in every respect than basic artificial fish-swarm algorithm.

The multirow improved is that overall artificial fish-swarm algorithm step is as follows:

(1) initialization operation is carried out to parameter, initialization Artificial Fish population number, the artificial and parameter such as initial position, Artificial Fish visual field Visual, Artificial Fish step-length Step, maximum iteration time, maximum attempts Try_number, the crowding factor;

(2) fitness of every bar Artificial Fish is calculated, and will there is the Artificial Fish state recording of optimal objective function to bulletin board;

(3) behavior that every bar Artificial Fish will carry out is evaluated, selects the most suitable behavior campaign of Artificial Fish instantly, have foraging behavior, collective behaviour, behavior of knocking into the back, behavior of eating, jump behavior;

(4) behavior selected by Artificial Fish execution, and upgrade the status information of Artificial Fish;

(5) state with the Artificial Fish of global optimum is upgraded;

(6) algorithm cycling condition is evaluated, if meet termination condition, then export optimal value, terminate program and run, otherwise jump to (2) and again circulate, until algorithm termination condition is satisfied.

Its algorithm flow chart as shown in Figure 3, the multirow of application enhancements is adaptive ability and convergence precision that overall fish-swarm algorithm improves algorithm, accelerate the convergence of algorithm time, and algorithm hide local extremum search global optimum in have excellent performance, apply the data that this algorithm can obtain all breaking up points of plate.

In addition to the implementation, the present invention can also have other embodiments, and all employings are equal to the technical scheme of replacement or equivalent transformation formation, all drop in the protection domain of application claims.

Claims (5)

1., based on a flame forming plate automatic forming method for Doubly curved surfaces, it is characterized in that, comprise the following steps:

1) importing of real deck of boat Tribon data;

2) the method Developing Surface of Doubly curved surfaces is applied;

3) the crack place of plate launched is carried out to the calculating of amount of contraction, utilize inference machine computing system to calculate and provide heater wire paths arrangement;

4) according to the heater wire arranged, system of processing selects the machined parameters that can reach designing requirement inside flame forming plate expert system, determine initial heating position by the heater wire path provided, drive heating head and water-cooling apparatus to realize the robotization processing of flame forming plate;

5) first after completion of processing, treat that steel plate is cooled to normal temperature, by camera filming apparatus and image processing system module, matching is carried out to the curve form of the plate after processing;

6) on the basis of matching, again One step development is carried out to the plate after processing, after expansion, calculates amount of contraction, and with initial deployment amount of contraction carries out difference comparsion, if amount of contraction does not reach designing requirement, and outside permissible error scope, then forward step 3 to); If the error between them within the range of permission, terminate processing.

2. as claimed in claim 1 based on the flame forming plate automatic forming method of Doubly curved surfaces, it is characterized in that, described step 4) in flame forming plate expert system be by a large amount of production experience knowledge sum up, the technological parameter of the processing of the common deck of boat is integrated and optimizes; The reliability model utilizing finite element emulation software to set up processes the emulated data that flame forming plate forming process is carried out and optimizes; One that a sets up expert system knowledge base in conjunction with the deflection of working process parameter, deck of boat oneself factor, environmental factor and most soleplate.

3., as claimed in claim 1 based on the flame forming plate automatic forming method of Doubly curved surfaces, it is characterized in that, described step 2) curved surface unwinding method be:

The expression carrying out uniform B-Spline to the ship hull plate imported carrys out the curved surface of this real deck of boat of matching, minimum principle is changed based on keeping the distance launched on the curved surface of front and back between all net points, initial delta expansion is carried out to curved surface, using initial expanding data as iteration initial value, be the data that overall artificial fish-swarm algorithm carries out that iterative obtains all breaking up points with the multirow improved.

4., as claimed in claim 3 based on the flame forming plate automatic forming method of Doubly curved surfaces, it is characterized in that, described curved surface unwinding method is:

That after keeping the distance in adjacent mesh in the pre-deployed, variable quantity is minimum as far as possible based on extreme value thinking method of deploying cardinal rule, point M (x, y, z) be the point on original curved surface, its data can calculate through fitting, point m (x, y) for launching the rear number of planes strong point corresponding with it, the expansion of curved surface is according to raw data points M (x, y, z) calculate the process of m (x, y), then approximate expansion problem can be grouped into following extreme-value problem:

$s=\frac{1}{2}\underset{\mathrm{ij}}{\mathrm{\Σ}}\underset{N}{\mathrm{\Σ}}{(L_{\mathrm{ij}}-l_{\mathrm{ij}})}^2=\min ---\left(1\right)$

In formula: L _ij=| M _im _j| be the distance of its original curved surface any two points;

L _ij=| m _im _j| for launching the distance of latter corresponding 2;

N is adjacent with M (x, y, z) coordinate a little;

To a M (x, y, z), N is the coordinate of eight points around it, the distance L of some M (x, y, z) and adjacent eight points before namely launching _ij, with the distance l putting m (x, y) and adjacent eight points after expansion _ij, variable quantity is minimum;

This equation is about variable x _ijand y _ijnonlinear equation, solve the coordinate that this equation can obtain launching rear each point.Can obtain equation analysis, ask the necessary condition of s minimum value to be that s is to x _ijand y _ijpartial derivative be 0, that is:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=0\end{array}\right.---\left(2\right)$

Respectively to x _ijand y _ijdifferentiate, above formula becomes:

$\left\{\begin{array}{c}\frac{\∂s}{\∂x_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](x_{\mathrm{ij}}-x_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\\ \frac{\∂s}{\∂y_{\mathrm{ij}}}=-\underset{N}{\mathrm{\Σ}}[{(x_{\mathrm{ij}}-x_{\mathrm{ab}})}^2+{(y_{\mathrm{ij}}-y_{\mathrm{ab}})}^2-{L_{\mathrm{ij}}}^2](y_{\mathrm{ij}}-y_{\mathrm{ab}})/{l_{\mathrm{ij}}}^2=0\end{array}\right.---\left(3\right)$

In formula, ab is the coordinate of some M (x, y, z) consecutive point.

5., as claimed in claim 3 based on the flame forming plate automatic forming method of Doubly curved surfaces, it is characterized in that, the multirow of described improvement is that overall artificial fish-swarm algorithm step is as follows:

(1) initialization operation is carried out to parameter, initialization Artificial Fish population number, artificial and initial position, Artificial Fish visual field Visual, Artificial Fish step-length Step, maximum iteration time, maximum attempts Try_number, crowding factor parameter;

(2) fitness of every bar Artificial Fish is calculated, and will there is the Artificial Fish state recording of optimal objective function to bulletin board;

(3) behavior that every bar Artificial Fish will carry out is evaluated, select the most suitable behavior campaign of Artificial Fish instantly;

(4) behavior selected by Artificial Fish execution, and upgrade the status information of Artificial Fish;

(5) state with the Artificial Fish of global optimum is upgraded;

(6) algorithm cycling condition is evaluated, if meet termination condition, then export optimal value, terminate program and run, otherwise jump to step (2) and again circulate, until algorithm termination condition is satisfied.