M.Sc. Thesis

Goran Majstrović, Load Balancing in Distribution Network by Means of Fuzzy Logic Approach, M.Sc. Thesis, University of Zagreb, Faculty of Electrical Engineering and Computing, Departement of Power Systems, Zagreb, Croatia, September 2001

Despcription: Electricity market opening introduces a lot of new possibilities in different aspects of organization, planning and control of power system. New shareholder relations and presence of competition define new rules in market environment. Power supplier is forced to adopt new, more precise and convinient measure procedures and generally to improve security and quality of supply. Consequently, in market conditions all system components are used maximally before new investment takes place. All companies in market environment are based on economic performance indicators and profit maximization. Simultanously, demand growth and new line routes problems are taking place in almost all power systems.

According to other system experiences, especially United States, Japan, Korea and India, it can be predicted that load balancing problem in distributino networks will increase and become important issue on different operating sessions and scientific conferences. Obviously, new efficient methods of load balancing in limited capacity distribution system is needed.

The main taks of this M.Sc. Thesis was to define method for optimal switching procedure in radial networks in order to balance the load in the system. In that way potential overload risk can be significantly reduced. At the same time it is necessary to balance phase load, too. Load balancing can be preventive and corrective. Preventive load balancing assumes procedures before thermal limits are reached, while corrective load balancing assumes solving of overload problem. In normal system conditions this method can be used for both types of load balancing.

Method defined in this paper is based on fuzzy set theory. Object function is defined in a way to equalize load index of all system element with system load index. The problem becomes more complex if the element is loaded assymetrically. Also, all switching device lifetimes are strictly defined by total number of switching operations. So, whole balancing procedure should be realized by minimal number of switching operations. In that way we have complex optimizing problem with three independent objective functions: system load balancing, phase load balancing and minimization of number of switching operations. And that is the main reason why fuzzy set approach was chosen for this multifunctional optimization problem.

Aoki's load balancing method was also introduced in the paper and it is compared to developed fuzzy logic method.

Fuzzy logic method is applied on real distribution network of city of Pula, Croatia and it showed excellent results.