109 8 th International Symposium „Topical Problems in the Field of Electrical and Power Engineering“ Pärnu, Estonia, January 11-16, 2010 Algorithm for Multi-Stage Scheduling in Intelligent Electric Transport System Ivars Alps, Mikhail Gorobetz, Anatoly Levchenkov Riga Technical University mgorobetz@gmail.com , ivars.alps@rtu.lv , levas@gmail.com . Abstract This paper describes mathematical formulation of multi-stage scheduling task and algorithm for its solution using artificial immune system in intelligent electric transport system. The task of scheduling in transport system is represented as one of the most well-known flow shop problem. Artificial immune system finds conflicts in existing schedule and propose optimal schedule of electric transport by waiting time and electric energy consumption criteria. As numerical example public transport schedule in Riga is proposed to be optimised. 1. Introduction Nowadays in metropolis number of vehicles is increasing day by day. It is a reason for a lot of problems for public transport, causing traffic jams, schedule violation, etc. The tendency is to combine different transport types such as tram, trolleybus and bus in one stop. In city centre it causes scheduling conflicts in peak hours. This is the cause for electric power overconsumption and public transport delay. To provide quick, cheap and comfortable passengers delivery with public electric transport in metropolis there was and will be topical problem. On the one hand, the amount of vehicles in the streets and electricity expenses are going up day by day, traffic jams become from bad to worse, public transport stays in traffic jams longer time and use electricity increasingly, but on the other hand the citizens require public transport as fast and as cheap as possible. Optimal traffic organization, coordination of traffic lights operation with transport moving on the routes according to an optimal schedule will help to solve this actual problem for any big city worldwide [6]. This paper describes a new application of artificial immune algorithm, to create schedule for optimal electric transport flow on routes. In terms of artificial immune systems, multi-criteria target function is defined as antigen. Schedules which are the solutions of the problem are defined as a set of antibodies. Each antibody represents a sequence of stages to perform all operations of each vehicle on public transport route. Randomly generated schedule is evaluated according to the target affinity function. As genetic based methods, artificial immune systems create new population of antibodies using specific procedures as diversification, clonal proliferation and hypermutation, which imitate features of biological immune system [10]. Processing continues until the solution is found or predefined stopping criteria is achieved. 2. Problem Formulation To organize an optimal electric transport units schedule on crossroads is very important for standby time decreasing and prevent the dangerous situations. Currently in inspected part of route transport units motion schedule is create without regard some unexpected delaying in peak on time and so on. They have not shared control system and have not coordinated schedule along public transport routes. This paper proposes to use immune algorithm for multi-stage scheduling as solution for traffic jams eliminating in the city. Route is divided into several separate nodes – stages ad thereby each section of route are examine as one stage in multi-stage task. If electric transport unit’s motion is coordinated according to optimal schedule, this could decrease electric transport standby time, eliminate often breaking and acceleration, finally save electricity and prevent dangerous situations. Main task of research is to create programmable model for transport unit motion in selected part of Riga public transport routes. Main goal of this research is to find out optimal motion schedule used immune algorithm. Fig. 1. Transport motion organization block diagram