Algorithm Component Package Development For Urban Power Network Planning And Identification Method On Connection Modes Of Medium Voltage Cable Network

As a powerful assistant decision-making tool and working platform, urban power network planning software can effectively reduce the work of planning and improve work efficiency. And algorithm package is of vital importance to urban power network planning software for composing the core part of planning software.At present, most research and development on algorithm for urban power network planning were confined to algorithm model in a certain aspect instead of design and implementation of algorithm in an integrative manner. A feasible scheme of design and realization of algorithm in form of component package is introduced in this paper. First, based on unified data model for power network and whole design of main algorithms for urban power network planning, algorithms are divided into four layers, which are network topology layer, power network topology layer, basic application layer and advanced application layer. Then, according to planning process and its features, the algorithm programs are organized and embedded into a computer decision-making support system for urban power network planning in form of component package. The proposed feasible scheme can reduce the algorithm and system coupling and improve reusability, extensibility, portability of algorithm. Besides, the algorithm component package cannot only effectively support the development for urban power network planning system, but also provide commercial algorithm component for other electric network analysis system.Network connection modes identification is one of the most basic tools for distribution network planning and evaluation. A program for connection modes identification of medium voltage overhead network is implemented via the approach presented by literature [4]. However, with the fast development of cable network in urban power network, connection modes identification of cable network is becoming a challenge to network planning for its increasing complex structure. In order to solve this problem, an improved pattern recognition based approach suitable for both overhead and cable network is proposed and an algorithm program is implemented. By taking feeders and power consumers as the basic research object and introducing the concept of group of network feeders and power consumers, connection modes of cable network can be identified effectively. The proposed approach has been successfully applied in a simple case and a real urban power network. The result verifies that proposed method is solid and correct.The developed algorithm component packages for network planning have been successfully implemented in a system computer decision-making support system for urban power distribution network planning (CNP4.0) by Tianjin University. And the system has been successfully applied in Nanjing, Shanghai, Guangzhou, Xiamen and other cities in China.