Modified Particle Swarm Optimization Algorithm Applied In Power System Reactive Power Optimization

In modern power systems, power quality is the key to ensure the safe and stable operation of the entire system. Voltage quality is an important indicator in measuring power quality. Reactive power is an important factor in influencing the voltage quality, therefore, it is necessary for the system to make reasonable reactive power planning. Reasonable reactive power optimization will not only improves the reactive power of the entire distribution effectively, but also reduces the active power loss of the system effectually.Reactive power optimization problem of the power system is a very complicated nonlinear programming problem. It contains a number of variables and constraints, and continuous variables and discrete variables coexist in the variables, so it needs appropriate solutions. Equality constraints in the constraints namely flow equations is a nonconvex high order equations, it also contains multiple variables and constraints, it is very difficult for solving if we adopt general mathematical method, it needs to adopt the reasonable calculation method to solve. In reactive power optimization, it is also need to select appropriate optimization algorithm, to choose an appropriate algorithm as the optimization method after having a comprehensive analysis of all the traditional optimization algorithms and artificial intelligence algorithm’s characteristics.Based on the features of reactive power optimization, this paper selects the Particle Swarm Optimization Algorithm as the optimization method, uses P-Q decomposition method for power flow calculation and selects the active power loss minimum value as the objective function which also combines penalty function for a model. The paper introduces the basic principles and implementation of the particle swarm optimization algorithm process in detail, states various improvement measures for the particle swarm optimization algorithm. Based on particle swarm optimization algorithm’s premature and easying to fall into local optimum and other shortcomings, the paper gives some corresponding improving measures, it uses the grouping variable inertia weight, divides the population into two size groups, that is large group with a typical linear decreasing strategy and small group with a nonlinear decreasing strategy based on arctan function; at the same time it selects the linear strategy for acceleration factor namely C1is linear decreasing, C2is linear increasing; it also combined with the choice operation in the genetic algorithm, through these measures, it effectively improves the performance of the algorithm, makes the algorithm jump out of local optimal solution, searches the region of the global optimum value, and converges the value to the global optimal value. In order to make up for defects that the selection operation brings the loss of the population diversity, this paper adds a disturbance factor, ranks each particle ’s fitness function values in order, and the most fitness function value which are much better are added a small perturbation factor, in order to increase the diversity of the population, for the small particles whose fitness function value are worst are taken to initialization operation, it solves the problem of the loss of diversity effectively.The Modified Particle Swarm Optimization Algorithm (MPSO) is applied in reactive power optimization, the paper emphatically introduces some problems on the application, such as dispose of the discrete variables, determination of the fitness function, judgment on the terminate conditions of the algorithm, and selection for the power flow calculation method. For the discrete variables, the paper uses integer and real hybrid coding strategy, this strategy has many advantages, such as no error, high accuracy and so on. This paper gives calculation steps and flowchart of the Modified Particle Swarm Optimization Algorithm for reactive power optimization. In order to verify the effectiveness and applicability, this paper finally uses the modified particle swarm optimization algorithm in IEEE-14and IEEE-30node standard test systems, uses MATLAB language for programming and simulation, and compares with the Standard Particle Swarm Optimization (SPSO).Through the analysis and comparison of the simulation results, it verifies its effectiveness, using modified particle swarm optimization algorithm can effectively reduce the system power loss and improve voltage value, compared with the standard particle swarm algorithm, the improved particle swarm algorithm has a better global convergence performance and a faster convergence speed.