Modeling and Optimization for Cost Optimization Model of Material Distribution System in Power Distribution Network--A Case in Hunan

This paper proposes a method to address several issues in the distribution system of distribution network materials, including non-transparent distribution costs, low punctuality of distribution, and frequent misreporting and misreporting of distribution costs by local warehouses. The method involves constructing a distribution hierarchy for distribution network materials and developing a mathematical model to calculate the costs associated with distribution, loading and unloading, and warehouse finishing. Additionally, a node identification mechanism is established based on the demand for distribution network materials to accurately identify nodes that can be removed from distribution routes. Furthermore, a distribution route optimization algorithm is developed to optimize distribution costs. The effectiveness of the method is then verified through its application to the distribution of distribution materials for the Hunan State Grid Corporation.


Introduction
The distribution network is a crucial component of the power system as it transports electricity generated by power plants to end users.Materials of distribution network components such as transformers, switchgear, cables, and lines have great impact on the efficiency and stability of the power system.However, there are currently issues including high loss of goods, expensive distribution costs, and limited information technology integration in China's logistics and distribution of distribution network materials [1,2].For the distribution of emergency supplies related to electric power supplies, Liu et al. [3] proposed a three-level network post-disaster material dispatch distribution model and improved NSGA-II algorithm based on the application of big data with the goals of post-disaster relief material demand prediction, distribution path optimization, distribution node optimization, and disaster victim

Framework of distribution system for distribution network materials
Before constructing the mathematical model, we first analyzed the distribution system framework of State Grid Hunan Corporation.The framework consists of three layers: (1) The first layer is the finished goods depots, which distribute materials to the swing depots.
(2) The second layer includes 14 municipal swing depots responsible for receiving materials from the finished goods depots and delivering them to terminal depots in districts and counties under each prefecture.These swing depots can also send materials to other swing depots if needed.
(3) The third layer comprises terminal depots at the county level, which receive materials from swing depots and deliver them to local construction sites for power facility maintenance.They can also send materials to other terminal depots or swing depots when necessary.To simplify our analysis, we will focus on these three layers (swing depots, terminal depots, and construction sites) as shown in figure 1. Arrows indicate possible directions of distribution.

Analysis of old and new energy cost and carbon emission modeling
The cost of energy consumption is a critical factor in truck operations, calculated by multiplying the cost per mile with the distance traveled.Traditional trucks predominantly rely on diesel fuel as their energy source, while new energy trucks utilize electricity.The unit mileage cost represents the energy consumed by the truck during its journey.For conventional trucks, diesel fuel serves as the primary energy source, and its price fluctuates based on market dynamics and international oil prices.
Consequently, the unit mileage cost includes the unit price of diesel along with associated taxes and fees.The energy cost calculation model is outlined below: From an environmental and sustainability standpoint, the Ministry of Industry and Information Technology has outlined key objectives for automobile standardization in 2022.The carbon emission calculation model is as follows:

Analysis of distribution cost model of distribution network materials
Taking Hunan State Grid Company as a case study, the optimization model for accurate distribution of materials within the distribution network can be divided into three main modules: distribution cost and loading/unloading, inventory finishing cost, and the value of cargo damage resulting from traffic accidents.

Distribution and transportation model
The State Grid Hunan Company uses land transportation for distributing network materials.The transportation vehicles used are general freight cars, which can be categorized into five types as shown in Table 1.The order information includes the distribution cost (Wq), the issuing point (i) and receiving point (j) of the material, order number, and line numbers corresponding to each order.Each order may have one or more line numbers associated with it.
The optimized model for efficient distribution of power grid materials can be divided into three stages based on the starting and ending points.The first stage involves delivering goods from the finished product warehouse to either a transit warehouse or a terminal warehouse.The model that considers these scenarios is represented by equation ( 3): The second stage involves distributing goods between the swing depot, terminal depot, and swing depot.To design a model for this phase, four components need to be considered: fixed costs, variable costs for mileage overruns, costs for multiple distribution points, and limited costs for mileage overruns.
The third stage is the end of distribution, the construction site to the terminal depot or swing warehouse self-collection stage, taking into account the cost of transportation and the need for timeliness, this stage of distribution is often selected as the self-collection destination j from the warehouse with the smallest distance and the inventory containing the required distribution network materials, so that the distance does not exceed the limit mileage L, < :

Loading and unloading, warehouse organization cost model
The first part is to obtain the necessary cost by multiplying the tonnage of loading and unloading, sorting materials as a unit billing quantity of manpower and warehouse equipment loading and unloading, sorting quantity su and the unit price by: The second component is the truck crane rental cost billed by the working hours , which is obtained by multiplying the decision variable with the number of bench shifts and the unit price : The joint summation of equations ( 3)-( 7) leads to the following cost optimization model for distribution, loading, unloading and sorting of materials in the national distribution network based on big data:

Model Assumptions and Constraints
(1) Model assumptions:1.The difference between delivered goods is solely determined by the product of the weight factor and the corresponding quantity indicated in the delivery charge sheet number and document line number.2.It is assumed that the volume of all unloaded and sorted goods is smaller than the remaining volume capacity of the node.3.The operating units involved in loading, unloading, and sorting are considered identical without any variations.4.It is assumed that all loading and unloading units at each node can adequately fulfill the requirements of loading and unloading orders.
(2) Binding Conditions: , , The Equation ( 8) represents the State Grid accurate material distribution cost model, consisting of a fixed number of kilometers.Equation ( 9), Equation (10) table type order all distribution vehicles on the remaining amount of goods, the amount of transportation between the nodes to meet the requirements of the nodes on the distribution of vehicle restrictions;Equation (11) represents the remaining cargo volume of the distribution vehicles in the order after the end of the loading and unloading process at the node;Equation (12) indicates that the number of distribution vehicles responsible for each node warehouse in the order is greater than the number of requirements.Equation (13) and Equation ( 14) indicate that the distribution needs to meet the demand of each order, each demand point and the entire distribution network, respectively;Equation (15) and Equation ( 16) indicate that each demand point can be distributed by more than one vehicle.Equation (17) indicates that when traffic accidents and other special circumstances occur, the order delivery fee is calculated as 0, and only the value of cargo damage is calculated.

Example of application
The aforementioned results were obtained by executing orders in three distinct stages.By inputting 6,078 distribution orders and 400 loading and unloading orders from State Grid Hunan Corporation into the program, running the aforementioned steps repeatedly will generate the corresponding outcomes.A total of 6,077 transportation orders were examined in this study.The average distance covered per order was found to be 170 kilometers, resulting in a cumulative recorded mileage of 1,033,090 kilometers, equivalent to 10,330.9 hundred kilometers.For the purposes of this case study, we specifically chose the Wuling Xiaka as the conventional diesel truck model and the Dongfeng E30 as the new energy vehicle model.

4.conclusion
(1) This paper offers a comprehensive analysis of material distribution within the distribution network, encompassing aspects such as distribution, loading and unloading handling, and cargo loss expenditure.By abstracting and integrating these processes, an optimization model for the cost of material distribution in the distribution network is established, taking into account the interplay between these factors.
(2) The study conducts a thorough examination of the costs and carbon emissions associated with new energy trucks compared to traditional diesel trucks.It explores both energy costs and carbon emissions, highlighting the advantages of increasing the adoption of new energy vehicles.This recommendation holds particular relevance for Hunan State Grid Company, which heavily relies on traditional diesel trucks.
(3) By comparing the actual distribution costs and loading/unloading expenses with the results obtained from the corresponding procedures of the model, it is observed that the actual costs surpass those calculated by the theoretical model.Through analysis, it is identified that the material distribution within the distribution network of State Grid Corporation in Hunan Province faces challenges such as a low level of informatization, high loading/unloading handling costs, and significant management difficulties.

Figure 1 .
Figure 1.Distribution system framework diagram

Table 1 displays
the symbols and descriptions of each element in the mathematical model.

Table 1 .
Symbol Description Loading/unloading time at node i for unit u with loading/ unloading/ finishing unit category β in order s The rated volume of vehicle k of model α in distribution order p U Collection of all loading/unloading units within order s The remaining quantity of goods when vehicle k of model α in distribution order p leaves the node u Serial number of the loading /unloading/sorting unit in the set U Distance from node i to node j Energy costs Decision variable, vehicle k with model α in order p is shipped between nodes i,j as 1, otherwise 0 Energy consumption per 100 km

Table 2 .
Information on new and used energy vehicles for distribution of network materialsCalculating the above data with the total mileage, we can get the comparison of energy cost and carbon emission of new and old energy trucks responsible for distribution network materials distribution of State Grid Hunan Company from January to May, as shown in table3.

Table 3 .
Comparison of Energy Costs for Old and New Energy Vehicles