The purpose of research is to develop new mathematical models and algorithms for intelligent transport system for optimal electric rolling stock speed control to reduce acoustic emissions and consumption of electrical energy of the train. Authors proposed to use optimal speed control algorithm for intelligent transport control system, which solves multi-objective optimisation task by noise reduction and energy saving criteria without infringements of the schedule. Proposed algorithm takes in account also optimal energy consumption that is achieved by optimal and safe control of electric vehicle using intelligent devices. The algorithm of optimal speed control is provided for train's agent. The algorithm includes intelligent speed control procedure and negotiations between intelligent agents. The specific environment is developed by authors for the dynamic modelling of intelligent railway system for noise and energy saving control. Results of modelling of these algorithms and procedures show efficiency of proposed methods in railway task solutions and give the possibility to reduce consumption of electrical energy by 24% and reduction of acoustic emission by 9%.