orthogonal problem of packaging material and its cutting
genetic algorithms and neural networks
Proposed for consideration is a method for the construction of heuristic algorithms for finding an approximate solution to NP-complete problems in a polynomial timeframe, which attempts to model natural human reasoning when solving such problems in practice. In the example problem of the orthogonal packing of a container and cutting of materials in 2- and 3-dimensional cases, an algorithm for their approximate solutions, based on the use of heuristic functional quality (multi-heuristic) is considered. In addition, the work gives examples of possible methods of optimizing the coefficients in this algorithm using genetic algorithms and neural networks, as well as some features of its software implementation and deployment production.