Dissertation abstract

The problem of optimal use of raw material is very important for many industries. For the iron and steel industry enterprises this question becomes aggravated because of the crisis predicted for the nearest 10 years. Efficiency of saving of material resources in such situation is the major factor of efficiency increase of all enterprise production. The solution of this problem directly influences on the prime cost of production. Production wastes make up a significant part of the prime cost of output production so minimization of waste is a primary task.

One of the major factors providing saving of rolled metal by minimization of production wastes, is the rational stock cutting. Automation of cutting process is a very important part of this problem.

In the master’s work the control of the flying shears used for rolled metal cutting to lengths is considered. Flying shears are intended for moving with a high speed metal cutting. In many cases these machines working capacity defines rolling mill productivity.

In the master’s work the control of the flying shears used for rolled metal cutting to lengths is considered. Flying shears are intended for moving with a high speed metal cutting. In many cases these machines working capacity defines rolling mill productivity.

1. Finding of the optimum plan of rolled stock cutting at a continuous billet mill;
2. Estimation of length of a strip of metal in each case;
3. Formation and delivery of operating signal on the executive mechanism of flying shears.

Thus, functioning of system, which is developing, can be presented in the form of the following scheme:

At the first level the planning of metal cutting according to a package of orders is carried out. And it is considered, that the length of a metal strip for each kind of steel is constant. This is the problem of one-dimensional stock cutting, which could be solved by means of simple methods of linear programming.

However it becomes complicated because of incorporating additional conditions - metal strip cutting on "minus" over-measure, i.e. lower and upper bounds for ordered lengths is specified, and the plan of cutting should provide an optimum choice of sheet length in these limits.

There are two methods of this problem solution:

1. To solve the problem by means of methods of linear programming and to correct the solution by means of heurism.
2. To solve the problem by means of genetic algorithm.

At this step the time of the solution does not play the big role, therefore both methods of the solution are admissible.

At the second level, by means of sensors the current parameters of technological process are defined. Further probabilistic value of stock length is defined.

The problem arising at this level consists in drawing up of mathematical model for evaluation of metal stock lengths and in definition of necessary accuracy of technological process parameters measurement. The problem of this level should be solved on each step of the cutting (after each cut of shears).

At the third level, it is necessary to carry out correction of the cutting plan according to the correction of the value of stock length. Here the rigid demand in the form of restrictions on the time of the solution is made. This restriction is carried out at use of genetic algorithms at the solution of the optimal cutting stock problem.

At the fourth level, the control of flying shears is carried out directly. The main part of this problem is the synchronization knifes rotation speeds and metal movements speeds. The flying shears that carry out this technological process use the tachometer control system. The drives of shears work synchronously with a drive of a last mill stand considering adjustment signals from the operator and from the cutting machine. In the master’s work it is necessary to consider the ways of formation of these signals. The control of the cutting plan performance is carried out by means of the sensors that measure of output metal sheets lengths.

To the top

Note: presented autoreferat is not a final version of master's thesis autoreferat as the current research is in course and is planned to be completed by 31.12.2006.

What if you smile...