Research on differentiated classification system of distribution network based on grading and zoning

In order to cope with multiple pressures such as ensuring the safe operation of the distribution network, promoting market-oriented operation, and satisfying rigid investment needs, power grid companies must implement a lean management approach. In response to the lack of systematic classification and evaluation methods for the distribution network in precise investment and future planning, this paper proposes a multi-dimensional hierarchical sub-regional distribution network differentiation classification system covering provinces, cities (counties), urban power supply units, towns, and villages. In the dimensions of provinces and cities (counties), based on the data fitting analysis of per capita electricity consumption, the classification method of the development stage of the power grids is obtained, and the provinces and cities (counties) are classified into pioneering benefits, steady improvement, and excellence three stages; in the dimension of urban power supply units, based on the combination of "top-down" and "bottom-up", the urban power supply units are divided into three categories: A+, A, and B; in the dimensions of towns and villages, according to characteristics indicators and construction models, proposed classification schemes for towns and villages. Finally, the results of an example analysis show that the differentiated classification system proposed in this paper can achieve an accurate assessment of the current development status of the distribution network, and provide important guidance for the precise investment of the distribution network in the future.


I. INTRODUCTION
With the integrated development of energy revolution and digital revolution and the promotion of energy Internet in recent years, the development mode of distribution network is also very different, with diversified and personalized characteristics [1][2][3][4]. As an important part of power grid, distribution network is an important platform to connect transmission network and users and support demand side response management [5][6][7][8]. In order to meet the new requirements of economic, social and reform and development, and cope with multiple pressures such as ensuring the safe operation of power system, promoting market-oriented operation and meeting rigid investment demand, the rational planning and accurate control of distribution network project planning and construction is one of the basic conditions for whether the power network can supply high reliability and high quality power [9][10]. In view of the demand for lean management and accurate investment decision-making of distribution network, the differentiated classification and evaluation method of distribution network as the key link of its future development investment and decision-making [11][12] has gradually become a research hotspot carried out by scholars.
At present, experts and scholars have carried out relevant research on the classification system of distribution network, but there is still a lack of systematic and applicable evaluation methods for distribution network construction projects, which is difficult to effectively evaluate the standardization and effectiveness of distribution network project construction, which brings many uncertainties to the realization of distribution network construction scheme and expected planning objectives. At the same time, the distribution network form index system is huge, the coupling relationship between indicators is complex, which increases the difficulty of the whole classification process [13].
There are many methods in the classified evaluation of distribution network. The qualitative methods include analytic hierarchy process, fuzzy comprehensive evaluation

II. DISTRIBUTION NETWORK DIFFERENTIATION CLASSIFICATION ARCHITECTURE
The differentiated classification system architecture of the distribution network proposed in this paper is shown in Figure 1. The differentiated classification system of distribution networks proposed in this paper includes multiple dimensions of provinces, cities (counties), urban power supply units, towns and villages. In the dimension of provinces and cities (counties), the Logistic function is used to fit and analyze the per capita power consumption to obtain the classification results of the development stage of the province and each city's distribution networks. Based on the evaluation of distribution networks development stage and construction level, provinces and cities (counties) are classified into four stages: basic operation stage, development benefit stage, steady improvement stage and striving for excellent stage. In the dimension of urban power supply units, based on the combination of "topdown" and "bottom-up" methods, the urban power supply units are divided into three categories: A +, A and B, and the power supply areas in typical power supply modes of distribution network are subdivided into 28 types. In the dimension of villages and towns, through the multiobjective evaluation system, the towns are divided into five types: industry oriented type, commercial trade type, tourism development type, agriculture-led type, comprehensive type. The villages are divided into four types: industrial development type, leisure tourism type, Traditional farming type and livable comprehensive type.

A. DISTRIBUTION NETWORK DEVELOPMENT STAGE ASSESSMENT
The change of investment demand of distribution network is closely related to the development stage of power network. From the typical experience, the annual investment demand of power grid basically presents the situation of "low at both ends and high in the middle". The law of changes in electricity demand in developed countries such as Germany, Japan, and Britain after entering the advanced economic stage also confirms this law. The US electricity data collected by the Energy Information Administration over the years also proved the existence of this trend. This paper uses the Logistic growth model to analyze the change trend of the distribution network development level over time t, which is expressed as: Where L(t) is the development level of distribution network; c is the saturation value of distribution network development; a and b are parameters; t is time. From the graphical characteristics of the Logistic curve in Figure 2, it can be seen that the development of the distribution network presents a typical S-shaped characteristic, and its development stage is divided by the change of the curve tangent. The development speed and acceleration of the distribution network are both small in the initial stage. With the passage of time, the speed of development has gradually increased, and the acceleration has gradually increased. In the later stages of development, the construction of the distribution network has gradually become saturated, and the speed of development has gradually become flat. In this paper, the development of distribution network is divided into four stages, and evaluated from three aspects: development scale, technical level and benefit. The characteristics of each development stage of the distribution network and the investment priorities in different stages are shown in Table Ⅰ. The focus is on establishing a preliminary grid structure as soon as possible, and the development speed is slow.
The technical level of power grid is limited, which can not fully meet the requirements of distribution networks reliability.
Efficiency improvement is not its main focus, and there is a lossmaking operation.
Establish a preliminary grid structure.

Development benefit stage
Pay attention to the rapid expansion of the grid scale to meet the demand for electricity. Due to the development law of diminishing marginal costs, economies of scale appear.
The development of distribution network has accumulated a certain fixed cost, and the technical level has been improved to a certain extent.
The level of efficiency has initially improved, but given the characteristics of its public utility companies, it still cannot meet the requirements of marketization.
Rapidly expand the scale of the grid.

Steady improvement stage
The scale of distribution networks continues to expand, but the investment, operation, management and other aspects of power grid have disadvantages due to the rapid expansion of power grid scale, and the possibility of diseconomy of scale increases.
The operation, control and automation level of power grid is improving day by day, which basically guarantees the power supply reliability of distribution networks.
There is a development bottleneck in the improvement of power grid operating efficiency. Overcoming this bottleneck depends on the coordinated development of grid scale, technical level and management level.
Pay attention to the transformation of power grid development and overcome the bottlenecks that hinder the scale effect.
Striving for excellent stage The scale development tends to be stable, and the leap forward development of power grid depends on the emergence of a new round of major changes in energy technology.
With the continuous improvement of informatization, automation and intelligence of power grid, the speed of technology updating is accelerated, and the power supply reliability of distribution networks is fully met.
The distribution networks development strategy focuses on the coordination and improvement of multi-level benefits.
Improve the efficiency of distribution networks at multiple levels.

B. DISTRIBUTION NETWORK CONSTRUCTION LEVEL ASSESSMENT
In this paper, by constructing a two-dimensional classification model of "power consumption per capitapower grid development adaptability", the classification results are used to evaluate the development and construction level of power supply companies in various provinces and cities, and are used to guide the determination and allocation of future power grid development investment, and to clarify the investment of power supply companies in various provinces and cities. The structure of two-dimensional classification model of distribution network construction level is shown in Figure 3. In this paper, the selection indicators and specific analysis steps of the two-dimensional classification model of "power consumption per capita-adaptability to grid development" are as follows: (1) Electricity consumption per capita. The actual income is the product of the electricity sales and the power transmission and distribution price, and attention should be paid to the investment intensity of incremental investment in areas where the electricity stimulus effect is obvious. Therefore, development investment should be tilted towards regions where the growth of electricity is relatively stable or there is room for higher growth. The electricity consumption of each city is evaluated by two indicators: per capita electricity consumption and per capita electricity consumption. Comprehensive evaluation of per capita electricity consumption in various regions.
Where S is the comprehensive evaluation score of electricity consumption per capita; L E is the electricity consumption per capita; E is the electricity consumption per capita.
(2) Adaptability of distribution network development. The development status of the distribution network is an important indicator that reflects the investment needs of various regions. For regions where the grid is weak, especially the distribution network, which requires a large amount of capital investment, these regions should be tilted when the investment arrangements for the transformation and upgrading of the distribution network are made.
The paper analyzes the asset quality of distribution network from the following two aspects: distribution network status (index: capacity-load ratio, N-1 pass rate, interconnection rate), power supply quality (index: power supply reliability rate, voltage qualification rate). The evaluation index system of distribution network development status is shown in Figure 4. Comprehensive evaluation of the adaptability of regional power grid development, the main indicators include power grid status and power supply quality. The comprehensive evaluation scores of grid development adaptability are as follows: Where the power grid state is Gs, the power supply quality of the distribution network is Gq, Dclr is the capacity load ratio of the distribution network, Dpr is the N-1 passing rate, Dir is the interconnection rate of the distribution network, Dpsr is the power supply reliability rate of the distribution network, Dvqr is the voltage qualification rate of the distribution network, and ɑ1、ɑ2、 β1、β2、 β3、γ1、γ2 are the weight factors.

IV. DIFFERENTIAL CLASSIFICATION OF URBAN POWER SUPPLY UNITS
In this paper, according to the dimension differentiation classification of urban power supply units, the combination of "top-down" and "bottom-up" is adopted. "Top down" refers to the power supply area defined by the technical principles of distribution network, and takes several application scenarios set by typical power supply mode of power grid as reference. The "bottom-up" approach is based on the actual work of the provincial unit system planning, taking into account the results of the completed unit system division of power supply enterprises in each city and the actual situation of saturated load density in the province, and putting forward the demand for the division of power supply units. Through the combination of the two methods, differentiated power supply units in various power supply areas are formed.

A. POWER SUPPLY AREA DIVISION
The division of power supply regions is mainly based on the load density of the administrative level or planning level, and can also be determined with reference to other factors such as economic development, user importance, power consumption level, GDP, etc. The specific power supply region divisions in this article are shown in Table Ⅱ. Agricultural and pastoral areas Note 1: σ is the load density of the power supply area (MW/km 2 ) Note 2: The area of power supply area is generally not less than 5km 2 . Note 3: When calculating the load density, the 110 (66) kV dedicated line load and the area of invalid power supply such as mountains, gobi, deserts, waters, and forests should be deducted.
Based on the division of A+ ～E power supply areas, according to the current status and development positioning of the distribution network in different regions, different functional divisions (administrative districts, business districts, residential areas, development areas, high-tech areas, etc ). Considering the technical indicators such as the planned area load density, power supply reliability rate, comprehensive voltage qualification rate, transfer capacity, etc., determine the target grid structure suitable for regional development, select appropriate standardized facilities, and ensure the reliable access of different types of power sources and users. Form a typical power supply mode of the distribution network adapted to the development of different regions. The power supply area in the abovementioned typical power supply mode of the distribution network is further refined as shown in Table Ⅲ.

A. DIFFERENTIATED CLASSIFICATION OF TOWNS
(1) Town classification index selection principles. The principle of scientificity: the new type of urbanization selected should be considered to be based on traditional urbanization and fully reflect the scientific connotation and characteristics of the new urbanization process. Principle of completeness and hierarchy: First, it is necessary to fully consider the indicators that can reflect the development situation, requirements and key issues of new urbanization, and avoid errors, omissions and repetitions; second, it is necessary to be logical and clear in organization and hierarchy, so as to lay a solid foundation for completeness. Operability principle: the practicality of the indicators should be considered, and the indicators that are not easy to operate should be selected as suitable alternatives or discarded.
(2) Town differentiated classification index system. According to the characteristics of towns under different economic development models and the results of the research on the differentiated classification of township power grids, this paper selects some indicators with different types of characteristics for calculation, so as to complete the classification of different types of townships. The classification index system of township areas is shown in Table Ⅳ.  Considering the actual operation environment of the index system and the difficulty of the actual operation, this part will use a function to calculate the corresponding index calculation method. The specific implementation is to use the two points of a function to determine the characteristics of a line, and select the appropriate value as the full score of 100 points and the passing score of 60 points. The corresponding index calculation formula is given. Delphi method is mainly used to determine the weight, and the numerical calculation method and weight calculation results are shown in Table Ⅴ.  100 50 The proportion of residential electricity consumption and primary production electricity consumption (%) 100 50 0.4 (3) Differentiation classification method of towns. The towns differentiation classification is mainly to calculate the corresponding score of characteristic industry according to the relevant indicators of each characteristic town. The higher the score is, the higher the degree of characteristic industry is, indicating that the town is closer to this characteristic industry. According to the score of each industry, the type of towns can be judged.
In information theory, entropy is a measure of the degree of chaos in a system, while information is a measure of the degree of order. The absolute values of the two are equal, but the signs are opposite. In the process of calculation, the greater the degree of data dispersion, the smaller the information entropy, the greater the amount of information it provides, the greater the impact of the index on the comprehensive evaluation, and the greater its weight; On the contrary, the smaller the difference of each index value is, the greater the information entropy is, the smaller the amount of information it provides, the smaller the impact of the index on the evaluation results, and the smaller its weight should be.
According to the secondary classification factors of index classification in Table Ⅴ, the classification indexes of each town are calculated.
Calculate the entropy value of the classification index c1-c4. When the entropy value is greater than 0.9, the index has no outstanding characteristics, and the town is considered to be a comprehensive type: When the entropy value is less than 0.9, the corresponding classification of Pi maximum value is taken as the classification of the evaluated towns. According to the calculation results, the differentiation classification results of towns distribution network proposed in this paper are shown in Table Ⅵ.

Industry characteristics Electricity consumption characteristics
Industry oriented type With industry as the leading industry, the GDP of the secondary industry accounts for more than 60% of the total GDP of towns The proportion of industrial power consumption to total power consumption is more than 50%, the load of electricity is high, the seasonal fluctuation is small, and the requirements for voltage quality and power supply reliability are high

Commercial trade type
With commerce as the leading industry, the GDP of the tertiary industry accounts for more than 60% of the total GDP of villages and towns The commercial and residential electricity consumption accounts for more than 50% of the total electricity consumption. The seasonal fluctuation of power load is small, and the peak valley difference between day and night is obvious. The requirements for voltage quality and power supply reliability are high

Tourism development type
With tourism as the leading industry, the GDP of the tertiary industry accounts for more than 50% of the total GDP of villages and towns Tourism and residential electricity consumption account for more than 50% of the total electricity consumption. The seasonal characteristics of power load are obvious. The power load in tourism peak season is high, and the requirements for voltage quality and power supply reliability are high

Agriculture-led type
With planting and breeding as the leading industries, the GDP of the primary industry accounts for more than 50% of the total urban GDP The electricity consumption of agriculture and animal husbandry and residents' living accounts for more than 50% of the total electricity consumption. The electricity load is greatly affected by the climate, environment, production cycle and other factors. The electricity consumption is large in key periods such as irrigation and drainage, harvest and cold, and the requirements for voltage quality and power supply reliability are high

Comprehensive type
Towns other than industry oriented, tourism oriented, commercial and trade oriented, and agriculture and animal husbandry oriented The characteristics of industrial power consumption are not prominent, and there are high requirements for voltage quality and power supply reliability

B. DIFFERENTIATED CLASSIFICATION OF VILLAGES
According to the differentiation of construction models, this paper divides rural areas into four types: industrial development villages, leisure tourism villages, traditional agriculture and animal husbandry villages, and livable comprehensive villages. The specific development characteristics and electricity consumption characteristics are shown in Table Ⅶ. Driven by the development of township industry, villages adopt the cluster development of rural industries, encourage enterprises to invest in the countryside by integrating local resource elements in rural areas, and vigorously develop modern rural industries, forming a block of "one village, one industry" and "one village, one product".
With rural industry and household small processing electricity load, the electricity load is high and seasonal fluctuations are small, with industrial electricity consumption accounting for more than 50%.

Leisure tourism type
In villages with superior natural conditions and abundant water resources, as well as villages with cultural landscape features such as ancient villages, ancient buildings and ancient dwellings, rural tourism, accommodation, catering, leisure and entertainment are developed. The villages, which are not rich in rural tourism resources, material and cultural resources, have complete facilities, The load of commercial operations, such as catering and accommodation, sightseeing and excursions, is dominant, with a sudden increase in load during the tourist season and significant changes in seasonal load, with the volume of tourism and residential life accounting for convenient transportation, and are close to cities, are suitable for leisure and vacation. more than 50%.

Traditional farming type
The main focus is on the development of agricultural crops and farm animal production, the realization of agricultural and livestock production aggregation, agricultural and livestock scale operation, relatively complete agricultural infrastructure such as farmland water conservancy, high rate of commercialization of agricultural products and agricultural mechanization, and abundant arable land resources per capita.
With modern agriculture, breeding industry, agricultural products processing electricity load, electricity load is low, agricultural sowing, harvesting and drainage and irrigation and other special time load surges, seasonal load changes are large, residential electricity consumption and electricity consumption of the first production accounted for more than 50%.
Livable comprehen sive type In villages with a larger number of people, larger scale, more concentrated residence and less prominent rural industrial development, we focus on solving the problems of poor rural living conditions, lagging infrastructure construction and serious environmental pollution to create habitable and environmentally beautiful villages.
The electricity consumption is mainly for residential use, and the electricity consumption load is low, influenced by air-conditioners, electric heaters and other temperatureregulating loads, with prominent time loads, and the electricity consumption ratio of each industry and residential consumption is relatively average.

Ⅵ. CASE ANALYSIS
This paper selects X provinces containing 14 cities (numbered A-N) for example analysis to further verify the rationality of the differentiated classification method for distribution network constructed in this paper.

A. DATA RESEARCH
The electricity consumption of a region not only reflects the development level of the power grid in the region, but also reflects its economic and social development level. The per capita electricity consumption further excludes the influence of the scale of electricity consumption, which is more comparable. Therefore, existing studies generally use per capita electricity consumption as the main basis to study the characteristics of the development stage of the power grid. This paper summarizes the per capita electricity consumption statistics of X province, and obtains the per capita electricity consumption of various cities in X province from 2000 to 2019, as shown in Figure 5. It can be seen from Figure 5 that the per capita power consumption of all cities in the X province is basically in a steady growth trend, and the per capita power consumption of some prefectures and cities has experienced a short-term decline and then resumed an upward trend. As of the end of 2019, the per capita electricity consumption in City A was the province's highest at 6965.87 kWh, and the per capita electricity consumption in City E was the province's lowest at 1504.42 kWh.

B. RESEARCH AND JUDGMENT RESULTS OF DEVELOPMENT STAGE
In this paper, the logistic curve model fitting prediction is carried out based on the electricity consumption per capita of each city in X province, and the development stage of the power grid of each city in X province is studied, and the corresponding development investment trend is analyzed. The result of the fitted curve is shown in Figure 6.

C. RESEARCH AND JUDGMENT RESULTS OF CONSTRUCTION LEVEL
Through the calculation and sorting of capacity load ratio, N-1 passing rate, interconnection rate, power supply reliability rate and voltage qualification rate of local cities in X province, the basic data onto distribution network development in X province is obtained, as shown in Table  Ⅷ. According to the two-dimensional classification model of "power consumption per capita-adaptability to grid development" proposed in this paper, the construction level of the distribution network and the future investment positioning of the cities in X province are studied and judged. The specific evaluation results are shown in Figure  8. It can be seen from Figure 8 that the comprehensive evaluation of per capita electricity consumption in cities A, C, and J is relatively high and the grid development adaptability is relatively high, while the comprehensive evaluation of per capita electricity consumption in cities I and F is low, and the grid development adaptability is relatively high. , City D, City E, City G, City K, City L, City N have low comprehensive evaluation of per capita electricity consumption and low grid development adaptability, while cities H and M have a higher comprehensive evaluation of per capita electricity consumption and relatively high grid development adaptability low. ability low.

D. COMPREHENSIVE EVALUATION RESULTS
Comprehensive distribution grid development stage and construction level research and judgment, analysis to obtain the evaluation results of grid development positioning of cities in X province are shown in Table Ⅸ. Considering the electricity consumption situation and grid structure comprehensively, based on the twodimensional classification results of "per capita electricity consumption-grid development adaptability", it is clear that the investment positioning of power supply companies in various regions is as follows.
City A, City J, and City C belong to the first category of units, that is, the comprehensive evaluation of per capita electricity consumption is high, and the grid development adaptability is high. The overall investment positioning of this type of unit: should continue to maintain stable development, and it is recommended to try as much as possible meet the investment needs of such units.
Cities I and F belong to the second category of units, low comprehensive evaluation of per capita electricity consumption and high adaptation to grid development. The overall investment of such units is positioned: there is a certain margin space in the grid, and investment benefit analysis can be conducted to improve the input-output ratio of grid investment according to the load development demand and the forecast of increased supply of electricity.
Cities B, D, E, G, K, L and N belong to the third category of units. low comprehensive evaluation of per capita electricity consumption and low adaptation to grid development. The overall investment orientation of such units: solve the current problems of the grid and try to meet the investment needs of such units involving electricity security.
City H and City M belong to the fourth category, high per capita electricity consumption and low adaptation to grid development. The overall investment position of these units is relatively weak, with problems in power supply capacity and reliability, but relatively good investment efficiency. The investment needs should be met from the target grid structure, equipment level, technology application.

Ⅶ. CONCLUSIONS
According to the needs of accurate investment and future planning of distribution network, this paper carries out hierarchical and subregional differentiation classification and evaluation. Based on the multi-dimensional distribution network differentiation classification system architecture, the research and judgment methods of distribution network development stage and distribution network construction level are proposed in the dimensions of province, city (county), in the dimension of urban power supply unit, the typical power supply mode of distribution network is divided into 28 categories, in the dimension of township and village, a classification method based on construction mode is proposed. Through the example analysis and verification, it can be seen that the distribution network differentiation classification system proposed in this paper can accurately evaluate the current development status of distribution network, give the key direction of future planning and investment by integrating various indicators, and provide solutions for the systematic evaluation and classification of distribution network.
The next research focus of this paper is to consider the quantifiable indicators such as distribution network cleanliness and low carbon, and further improve the distribution network classification and evaluation system proposed in this paper.