The purpose of the paper is to present new mathematical models and new algorithms for mechatronic systems based on decision-making and intelligent multi-agent approaches. Paper presents the decision making task for optimal energy consumption control by electric vehicle using intelligent devices. The mathematical model for intelligent agents of electric transport system is offered for the task solution. The algorithm of optimal speed control is provided for vehicle's agent. The algorithm includes speed control procedure and negotiations between intelligent agents. Paper proposes the solution for decision-making task of coordination of intelligent electric transport units and intelligent electric traffic control units with a purpose to reduce energy consumption. The model of electric transport system is defined and the algorithm for electric vehicle motion control with negotiations between electric devices using intelligent agent approach is presented. Paper provides general solution of diagnostic task of the electric device. The decision support system based on intelligent agents is proposed to analyze a workability of public electric transport units. The base of multi-agent is neural network trained by back-propagation algorithm. Intelligent multi-agent provides possibility to detect problem and offer solution immediately. It is a chance to avoid superfluous charges connected with problem detection, fixation and consequences of the problem. Also the negotiation algorithms of intelligent multi-agent are proposed to make decisions in the tasks of optimal delivery scheduling and routing in geographically distributed supply chains. The workability of the proposed models and algorithms is tested in a specific simulation environment. As an example, a tram system with real parameters is taken for the optimal energy consumption task solution. General algorithm for diagnostics in intelligent agent systems gives possibility to detect the problem of electric vehicle immediately, to fix it in some cases without human intervention or inform all other participants about the problem. Main advantages of using negotiations in intelligent agent systems are the possibility to coordinate action of all participants of the complex processes and to realize multi-criteria decision-making in scheduling, energy consumption optimization and diagnostic tasks. The results of coordination follows cost reducing, profit increasing, idle time minimization and safety in case of emergency. The general problems of addition power supply solution are proposed in the article, in case, if district recourses are not enough for these customers. For the development and investigation of the software agents models of electric power supply and transport logistics a complex of mutually co-ordinated models is investigated in the article. The complex of mutually co-ordinated models, are used for achieving coals of customers and experts of power supply process. A practical application of intelligent agents for decision making task for route choice of the power distributing substations service is under consideration. Intelligent agents system based on the proposed models and algorithms is created using web-technologies, a database and the appropriate programming languages that allow to realize negotiation easy and effectively.