Journal of Power and Energy Engineering, 2014, 2, 457462 Published Online April 2014 in SciRes. http://www.scirp.org/journal/jpee http://dx.doi.org/10.4236/jpee.2014.24061 How to cite this paper: Zhang, C., et al. (2014) Optimal Resources Dispatching Technology of Distribution Network Rush Repairing. Journal of Power and Energy Engineering, 2, 457462. http://dx.doi.o rg/10.4236/jpee.2014.24061 Optimal Resources Dispatching Technology of Distribution Network RushRepairing Chao Zhang, Xinhe Chen, Xing Xiong, Jing Zhou, Wenbin Zhang State Grid Electric Power Research Institute, Beijing, China Email: zhangchao5@sgepri.sgcc.com.cn, chenxinhe@sgepri.sgcc.com.cn, xiongxing@sgepri.sgcc.com.cn, zhoujing_0419@126.com, zhangwenbin1@sgepri.sgcc.com.cn Received Dec emb er 2013 Abstract Confronted with the r equi rement of higher efficiency and higher quality of distribution network fault rushrepair, the subject addressed in this paper is the optimal resource dispatching issue of the distribution network rushrepair when single resource center cannot meet the emergent re source demands. A multiresource and multicenter dispatching model is established with the ob jective of “the shortest repair starttime” and “the least number of the repair centers”. The optimal and worst solutions of each objective are both obtained, and a “proximity degree method” is used to calculate the optimal resource dispatching plan. The feasibility of the proposed algorithm is il lustrated by an example of a distribution network fault. The proposed method provides a practical technique for efficiency improvement of fault rushrepair work of distribution network, and thus mostly abbreviates power recovery time and improves the management level of the distribution network. Keywords Distribution Network; RushRepairing; MultiObjective and MultiResource Dispatching; Proximity Degree Method 1. Introduction The power recovery efficiency of the power grid depends on the matching degree between the damage of the power grid and the rushrepair capability. The rushrepair capability of the power gird depends on the repair re source reserves, field circumstance and the dispatching capability. The resource reserves are fundamental to the rushrepair work. And the resources dispatching technology is critical for the fast recovery to meet with the re quir ement of the distribution network fault rushrepair [1][5]. When the power grid is facing with the continuous or extensive attack, such as severe weather or geological disaster, rushrepair crew will be very busy and the repairing resource might not be able to meet with the emer gency resource demands. In a rushrepair work, when the repair resources are insufficient in local resource center, the decisionmaker relies on his previous experience to command. It might be very difficult for them to make an optimal decision under the condition that there are various kinds of resource demands form distributed resource centers, which
C. Zhang et al. implies that it is hard to make an optimal decision. Besides, there is not evaluation criterion for the decision made by the commander until the rushrepair work is done. In this paper, a reasonable evaluation criterion and corresponding techniques for the decision making are studied to figure out the optimal resource dispatching plan. 2. Modelling for Resource Dispatching 2.1. The Resource Dispatching Process When a power outage event happen in the power grid, the rush repairing command center analyzes the outage cause based on the information from DSCADA, electricity usage data acquisition system, 95,598 customer ser vice system, and etc.. According to the fault analysis, the command center recognizes the needed type and amount of the repair resources and then sends dispatching order to the resource centers. The dispatching process is shown in Figure 1. 2.2. Resource Dispatching Modelling Sufficient repair resources are the fundamental requirement for the distribution network rushrepairing. In a rushrepair work, there will be n involved resource centers, which are notated as . Meanwhile, there will be m (m > 1) types of repair resources needed in fault location A. The types of resources are notated as , and the amount needed for every resource is , respectively. and is as sumed to be the resource reserve and supplying amount of the jth type of resource from the ith resource center, in which , and . The delivery time from to A is ( ), assuming . The optimal resource dispatching plan implies that the starttime of the rushrepairing work is shortest and the number of involved resource centers is minimized. Ass uming the optimal plan is , which is specified as (1) where, ()()( ) { } 112 2k k ddd dd d '' ' ,, ,,, , j jjj Ax AxAx ϕ = represents the rushrepairing plan for the jth resource. If the selected resource center from all of n centers satisfy the demand of the jth resource, we have , ( ) (2) where, are the supplying amount of the jth resource from the selected resource centers, sep arately. The set of all the dispatching plans is Ω . Outage Outage Resource demands met ? Resource demands met ? Y N Resource dispatching from available centers Resource dispatching from available centers Check the available resources in the centers near the fault Check the available resources in the centers near the fault Resource dispatching command Resource dispatching command Rushrepairing ends Rushrepairing ends Fault Analyzing Fault Analyzing Resource dispatching in the center Resource dispatching in the center Figure 1. Rushrepair process.
C. Zhang et al. Definition 1: The number of the selected resource centers is and the starttime of the rushrepair work is , which implies that all the needed repair resources arrive at the fault location in with their demand satisfied. Thus, , (3) The objective of the resource dispatching is to minimize the starttime of the rushrepair work and the number of the resource centers involved. That is min( ()) min(( )) . T N st ϕ ϕ ϕ ∈Ω (4) The Equation (4) is a multitarget decisionmaking problem which can be solved by the technique of order preference by similarity to ideal solution. Therefore, the positive distance and the negative distance to the ideal solution can be worked out from the two following objective functions. (5) (6) Assume is the best and worst solution of the Equation (5), separately. And is the best and worst solution of the Equation (6), separately. The proximity degree between a dispatching plan and the best solution can be expressed as: 12 () () () () vvv NT RNT ϕϕ ωω ϕϕ ′ ′′ = + (7) Similarly, the proximity degree between the plan and the worst solution can be expressed as: 12 () () () () vv v NT rNT ϕϕ ωω ϕϕ = + ′ ′′ (8) In the Equation (8), and are the weight of “the number of the resource center” and “the starttime of the rushrepair work”, separately, and . The specific value can be obtained by specialists. In this pa per, both of them are 0.5. The relative proximity degree between the plan and the ideal solution can be ex pressed as follows: ， (9) Therefore, the multiobjective decisionmaking problem can be translated to the proximity degree problem between the suggested solution and the ideal solution. The solution which obtains the maximal proximity degree is optimal [6]. 3. The Solution to the Resource Dispatching Problem and should be solved firstly for the Equations (7)(9). 3.1. The Solution to the The resource centers involved should be near enough to make sure that the starttime of the rushrepair work is as earlier as possible. Assume the resource centers are ranked by the delivery time to the fault location from shortest to longest. If 1 00 jj qq pj jpj pp xx x − = = <≤ ∑∑ ,
C. Zhang et al. where , the optimal dispatching plan for the jth resource with the objective of the shortest delivery time can be expressed as: 1 11220 ( ,),(,),(,) j j q jjjq jpj p AxAxA xx ϕ − = ′=− ∑ …， (10) Thus, which is the delivery time of the jth resource from the resource center Aqj to the fault location, is the shortest delivery time for the jth resource. Similarly, the shortest delivery time for the other resources can be obtained. And the best solution to the Equ ation (5) is . Therefore 12 ()max( ,,,) m qq q Ttt t ϕ ′′ = (11) And (12) Notice that the longest delivery time is , that is: (13) 3.2. The Solution to the Similar to Section 3.1, assume the resource centers are ranked by the resource reserve of the jth resource from least to most. The optimal dispatching plan for the jth resource with the objective of the least number of the involved re source centers can be expressed as: 11 22 1 1 (,),(,), ,(,) j pj j p jk kjkkjkjkj i AxAxA xx ϕ − = = − ∑ … (14) where, is least number of the involved resource centers for the jth resource. Therefore, the least number of the involved resource centers of the rushrepair work is (15) And for every dispatching plan , (16) 3.3. The Solution to the Dispatching Problem In the resource dispatching optimization problem, not only the number of the involved resource centers is con sidered to be as small as possible, but also the starttime of the rushrepair work should be as early as possible. Therefore, the dispatching plan whose relative proximity degree is the biggest is taken as the optimal solu tion. The calculation steps of the are as follows: Step 1: Work out , ; Step 2: Let the set R = { } be the group of the repair centers, and the serial number y = 0; Step 3: Select combinations of centers from the set R. If there is not a combination feasible, go to step 9; Step 4: Let y = y + 1, Select the combination whose starttime is shortest from all of the availa ble feasible combinations; Step 5: Let , and if y = 1, let and ; Step 6: According to the Equation (7) and (8), calculate the proximity degree of between and the ideal solution; Step 7: Obtain the by Equation (9), and figure out the dispatching plan ;
C. Zhang et al. Step 8: Modify the set R by deleting the repair centers whose starttime is longer than ; Step 9: Let ; Step 10: If the length of the set R is larger than , go to step 3; or else, go to step 11; Step 11: Compare the obtained , and the largest one is optimal The flow chart of the algorithm is shown in Figure 2. 4. Case Study In this study case, the distribution power grid was affected by storm. Based on GIS technology, the rushrepair ing center confirmed that the fault location is A. The dispatching plan of the required repair resources should be optimized to ensure the efficient of the rushrepair work. In this rushrepairing, 32*insulator (XPW7)\25*pole (18 m) and 36*Cross Arm (1 meter) are required. There are eight available resource centers near fault location A, and their resource reserves are shown in Ta ble 1. The weights and are set to be 0.5. As is shown in Table 2, the biggest relative proximity degree is 0.6364 (y = 3). Therefore, the third dispatching plan whose starttime is 15 minutes is the optimal one. The delivery time for every center is 10, 12, 14, 15, 20, 22, 25 and 30 minutes, separately. Follow the algorithm proposed in this paper, the calculation results are shown in Table 2. The details of the optimal plan are shown in Table 3. Let According to (7) and (8), calculate the proximity degree of the solution to the best solution and the worst solution. Work out Let be the group of the repair centers , and let y=0, and Select the combinations of n’ centers from R Is there a combination feasible ? Is the length of the set R larger than Modify the set R by deleting the repair centers whose starttime is larger than According to (9), calculate the optimal index , and figure out the dispatching plan . Let y=y+1, and select the combination whose start time is shortest Compare the obtained , and the maximum one is optimal Figure 2. The flow chart of the resource dispatching.
C. Zhang et al. Table 1. The resource reserves. Resou r ce Insulator 12 5 10 9 9 6 24 7 Pole 10 8 4 12 4 15 10 5 Cross Arm 12 16 5 7 9 12 20 14 Table 2. Calculation results. 1 A1, A2, A3, A4, A5, A6, A7, A8 unfeasible 2 2 A1, A2, A3, A4, A5, A6, A7, A8 A1, A2, A7 (A1, 12), (A2, 5), (A7,15) (A1, 10), (A2, 8), (A7,7) (A1, 12), (A2, 16), (A7, 8) 3 25 0.5697 3 A1, A2, A3, A4, A5, A6 A1, A2, A3, A4 (A1, 12), (A2, 5), (A3, 10), (A4, 5) (A1, 10), (A2, 8), (A3, 4), (A4, 3) (A1, 12), (A2, 16), (A3, 2), (A4, 6) 4 15 0.6364 4 When the quantities of elements are smaller than n', the calculation ends. 5 Table 3. The optimal resource dispatching plan. Resou r ce A1 A2 A3 A4 Insulator (XPW7) 12 5 10 5 Pole (18 m) 10 8 4 3 Cross Arm (1 m) 12 16 5 6 5. Conclusion In this paper, the optimal resource dispatching issue of the distribution network rushrepair is studied. The dis patching model is established first with the objective of “the shortest repair starttime” and “the least number of the repair centers”. And a “proximity degree method” is used to calculate the optimal resource dispatching plan. The feasibility of the proposed algorithm is illustrated by an example of a distribution network fault. The pro posed method provides a practical technique for efficiency improvement of fault rushrepair work of distribution network, and thus mostly abbreviates power recovery time and improves the management level of the distribu tion network. References [1] Cui, W. and Wang, B.D. (20 02 ) Development and Application of an Electrical Rush Repair Scheduling Syst em. Auto mation of Electric Power Systems, 26, 6467. [2] Lu , Z. G., Sun , B. and Liu , Z.Z. (2011) A Rush Repair Strategy for Distribution Networks Based on Improved Discrete MultiObjective BBC Algorithm after Discretization. Automation of Electric Power Systems, 35, 5559. [3] Tao W.W., Zhang, H.B. and Ding, J.Y. (2010) Design of MultiLevel Synthetical Intelligent DecisionMaking Support System for Secure Operation of LargeScale Regional Power Network. Power System Technology, 34, 8086. [4] USCan ada Power System Outage Task Force (2004) Final Report on the August 14, 2003 Blackout in the United States and Canada: Causes and Recommendati ons . [5] Yang, Y.H. , Zhang, D.Y. and Ma, S. (2004) Study on the Architechture of Security and Defense System of Large Scale Power Grid. Power System Technology, 28, 2327. [6] Wan g, Y. and He, J.M. (2002) Reasearch on MultiResource Dispatch in Emergency System. Journal of Southeast University (Natural Science Edition), 32 .
