Solving feeder automation planning problem with mo

Solve feeder automation planning problems with 0-1 planning

classification number: tm715 document identification code: a

article number: (2000) optimal feeder automation planning using binary programming Wang Tian Hua,

(Beijing creative distribution automation D, Beijing 100085, China)

wang Ping Yang, FAN Ming-tian

（Electric Power Research Institute China,Beijing 100085，China）ABSTRACT：In this paper, an analytic reliability algorithm, in which the reliability indices such as ENS and PNS are expressed as polynomials of switch variables, is proposed firstly. Then, the feeder automation planning is formulated as an optimal switch placement problem. A new binary programming is employed to solve this problem. Two examples are used to demonstrate the feeder automation planning process.

KEY WORDS：feeder automation planning; energy not supplied (ENS); power not supplied (PNS); Binary progra let them know the application scope of materials mming1 introduction feeder automation is the key to improve the reliability of distribution power supply, and fault location, isolation and recovery are its main functions [1 ~ 3]. In order to ensure the reliability of power supply, traditional urban power distribution often adopts double circuit or even three circuit lines. Since the number of outgoing lines in the substation is increased, it is usually necessary to build a switching station to expand the number of outgoing lines. This substation switchyard user power supply mode not only has low equipment utilization and large line investment, but also increases the difficulty of outgoing line of substation [4]. Feeder automation technology only needs to install column switches, sectionalizers, reclosers and other equipment along the feeder, which can ensure the reliability of power supply through fault location, isolation and restoration of power supply functions, and reduce the maintenance and overhaul costs of power distribution

to realize feeder automation, we must first gradually transform the radiation power supply mode of the old city into the ring power supply mode, which has reached a consensus. However, how to segment the line and how to locate the ring switch are the problems that should be solved in the feeder automation planning stage [2,3]. Switch location is the most important decision variable in feeder automation planning. Once the switch location is determined, the communication mode and control strategy of distribution automation and the configuration of other equipment can be preliminarily determined

generally, the higher the reliability of the distribution system, the greater the investment. However, high reliability can reduce users' power failure losses and operating costs. The goal of feeder automation planning is to determine an optimal switch configuration scheme to minimize the total cost of reliability on the premise of meeting the reliability of important users

this paper first proposes a distribution reliability calculation method based on analytical method. The analytic method expresses the reliability indexes outage power loss (ENS) and outage power loss (PNS) as polynomials of switching variables, which can consider the impact of fault isolation and recovery on reliability. Then the feeder automation planning model is expressed as the optimal switch configuration problem, and the polynomial programming method is used to solve it. 2 distribution reliability algorithm based on analytical method 4 sampling quantity reliability is an important goal of distribution planning and design, and its indicators mainly include: ① system average outage frequency SAIFI; ② System average outage duration saidI; ③ Power loss ENS; ④ Power loss due to power failure, PNS, etc. Among them, SAIFI only reflects the reliability of the network structure itself, and saidI reflects the maintenance and management status of power companies. Ens and PNS comprehensively consider the frequency, duration and severity of the fault, and this paper focuses on these two "indicators. Assuming that the loss caused by power failure per kWh and kW to users is ckwh and ckw respectively, the loss cost of user power failure can be measured by the following formula: r = ens × CkWh＋PNS × Ckw (1) distribution reliability calculation is based on feeders. One feeder includes one main line and several branch lines. According to the size of the load, the number of users and the length of the line, the branch line can be divided into three categories. The branch of the first type is very short and the load is small. Generally, no protective equipment is installed. The branch of type 2 is short, and only fuses are installed. The third type of branch has long lines and heavy loads, and switchgear can be installed at the branch. In the calculation of distribution reliability, the first and second branches can be regarded as a part of the main line. In this way, the complex multi branch tree feeder can be decomposed into several typical feeders as shown in Figure 1. Figure 1 Typical feeder without class 3 branch

fig.1 a typical main feeder section2.1 case without class 3 branch

assume that the typical feeder shown in Figure 1 has n segments, and PI represents the sum of loads directly connected with segment I, such as P (1-2) = P3 + P4. Li、 λ i. Di and Xi (I = 1,..., n) respectively represent the length, failure rate, failure repair time and switching variables of segment I, Xi represents the configuration of the installed switch, xi = concept 1, changing the world width 0 to the installed switch, xi = 1 to the non installed switch, and DS represents the load transfer time. Y indicates whether to install the interconnection switch, y = 0 means to install the interconnection switch, y = 1 means not to install the interconnection switch

when branch I fails, the ENS index caused by isolated load is (2) when y = 0, the downstream load of branch I can resume power supply, and the outage time is ds; When y = 1, the power supply cannot be restored, and the power failure time is di. The ENS caused by these two conditions is