Research On Dynamic Reactive Power Optimization Of Distribution Network With Distributed Generations Based On Improved Differential Evolution Algorithm

The solutions about the dynamic reactive power optimization problem of the distribution network have a complex and nonlinear characteristics because of space-time coupling and decision variables as mixed integers.In the model,not only the dynamic change characteristics of the load but also the maximum action times of the control equipment in the optimized period should be taken into account.After the distributed generations(DG)is connected to the distribution network,the active power of DG is supplied to the distribution system,and a certain adjustment range of the reactive power is provided.Therefore,Taking advantage of the reactive power regulation capability of the distributed generations combined with the traditional reactive power regulation means has practical research significance.In this paper,the dynamic reactive power optimization problem of distribution network with distributed generations is studied based on the improved chaotic parameters and multi-strategy mutation algorithm.Firstly of all,the power flow algorithm with different types of distributed generations is improved,and the influence of distributed generations accessed in the distribution network is analyzed.The multi-scenario and the reactive power optimization mathematical model of distribution network with the minimum active network loss as the objective function is constructed in terms of the scenario occurrence probability,and the randomness of the distributed power output is eliminated.Secondly,a chaotic parameters and multi-strategy mutation differential evolution algorithm is proposed to solve the reactive power optimization problem.Compared with the traditional differential evolution algorithm,multi-strategy mutation increases the diversity and distribution of individual vector solutions,enhances the exploring ability of global optimal solutions.The chaos optimization theory is mixed into differential evolution algorithm for improving the inefficiency problem due to the initial value randomness,and strengthening the local depth search capability of the algorithm.And then,the mathematical model of dynamic reactive power optimization of the distribution network with the minimum all-day active power loss as the objective function is established based on the output combination scenario of distributed generations.The initial equipment action timetable is objectively formulated by the ordered clustering algorithm,the solving process of the dynamic reactive power optimization problem is simplified.Besides,the subjectivity of the general pre-action table method is avoided.Finally,there are some conclusions according to the simulation results analysis of a modified IEEE33 node distribution network system.The reactive power optimization method based on the full-scenario optimization can effectively adapt to the dynamic conditions of the distribution network accessed DGs.The purpose of reducing network loss and increasing voltage level is achieved.The Further considered dynamic reactive power optimization method including the timing change characteristics of load and the capacitor group switching times,it can significantly reduce the network loss and effectively avoid voltage over-limit,which is more realistic.The maximum number of switching times of the capacitor bank meet the constraint,and the number of switching times is reduced as much as possible by the application of the ordered clustering method.The simulation results also verify that the chaotic parameters and multi-strategy differential evolution algorithm proposed in this paper shows better convergence and stability in solving the problem,and the optimization performance is better.