In the process of transition to Industry 4.0, the importance of applying cutting-edge technologies such as machine learning and artificial intelligence to replace human operators in industrial processes is explained by the need to automate industrial production processes. Replacing qualified human experts with artificial neural networks opens up a lot of possibilities for the implementation of new methods of industrial process automation. The problem of industrial process automation is quite complex because the decision-making process of the human expert is accompanied by uncertainty. Artificial neural networks represent one of the basic branches of artificial intelligence. At the moment, they are used in various fields to solve problems for which classical methods are unable to provide practical solutions. Thus, the problem of developing and training artificial neural networks for solving industrial process automation problems acquires major importance in the design of artificial intelligence systems. The training process directly depends on the data set on the basis of which the neural network is designed.

The paper presents the results of research carried out to solve complex problems aimed at the efficient use of natural and energy resources. The objectives of the paper are achieved by identifying the control process based on a Multi-Agent system with distributed data processing that implements a Multi-objective optimal solution search model based on the application of a Genetic Algorithm with Collective Computation. The set of Agents presents a computational architecture that forms a structured network topology based on a P-Systems model presented in the form of a Venn diagram. The Object diagram and the Venn diagram of the P-Systems model are presented in the paper. The correctness of the developed models was verified on the basis of a control system of the artificial lighting process that provides for the minimization of energy consumption, while ensuring the quality of the lighting process.

The street advertising has undergone some significant changes in recent years: traditional billboards are gradually being replaced by electronic display devices (LED screens) that are able to change in real-time the broadcast advertising, thus allows the dynamic content management. This paper aims to develop an adaptive advertising strategy based on the preferences of the people in front of the screen. Each of them has a special application installed on their personal smartphone through which they can configure their interests regarding the broadcast advertising. These interest profiles are then collected by billboards which, based on them, select the most appropriate type of ad to run at that time. The proposed method focus on transformation of a simple display equipment into an intelligent one, capable of adapting the broadcast content to the requirements of the nearby audience and aims to maximize the efficiency of the billboard operation and at the same time bring maximum satisfaction to the target audience. The performance of the method was evaluated using agent based computer simulation.