Search

CN-122026321-A - Traction power supply system photovoltaic uncertainty and energy storage optimal configuration strategy considering regional difference

CN122026321ACN 122026321 ACN122026321 ACN 122026321ACN-122026321-A

Abstract

With the large-scale configuration and planning of Traction Power Supply Systems (TPSS) combining photovoltaic and energy storage, current research into photovoltaic uncertainty and photovoltaic energy storage capacity allocation requires large sample photovoltaic data. Under the influence of geographic differences and extreme climates, the sample data size is small, the photovoltaic fluctuation amplitude calculation is difficult to converge, and in most areas, the geographically compatible sample data is even more difficult to find for the capacity allocation planning of the Super Capacitor (SC). Challenges are presented to existing capacity allocation schemes and photovoltaic uncertainty solutions. In order to solve the problems of photovoltaic uncertainty and capacity allocation when no photovoltaic sample data or only short-term sample data are provided, a two-layer optimization model configuration based on a photovoltaic prediction model is provided. The optimization model is solved using a particle swarm algorithm embedded in a CPLEX solver. Finally, the economy and effectiveness of the model is verified by comparison with existing solutions.

Inventors

  • MA QIAN
  • LU YUFAN

Assignees

  • 湘潭大学

Dates

Publication Date
20260512
Application Date
20260121

Claims (5)

  1. 1. The traction power supply system photovoltaic uncertainty and energy storage optimal configuration strategy considering regional differences comprises the following steps: Obtaining geographical environment data and actual measurement load data of a traction substation and photovoltaic output data of a photovoltaic power station; building a photovoltaic prediction model based on geographic environment data to predict photovoltaic output; dividing the capacity configuration of the traction power supply system into an upper layer and a lower layer; establishing an upper layer variable model based on a photovoltaic prediction model and existing data for an upper layer; and for the lower layer, converting the upper and lower uncertain models into a determined model solution through interval linear programming.
  2. 2. The method for dividing capacity allocation of a traction power supply system into two layers according to claim 1, wherein according to the characteristics of photovoltaic output affected by weather and geographical environment factors, a photovoltaic prediction model is established by referring to regional and climate characteristics and other related parameters such as regional longitude and latitude, altitude, illumination angle, weather index and the like, and the objective function is as follows: Wherein the method comprises the steps of For the headroom direct irradiance, For weather type index, there are Wherein a is a constant, and generally 0.14 is taken; H represents the altitude of the place where the photovoltaic power station is located; The incident angle of sunlight at the time t; Taking a weather index value of the ith day, taking a value according to weather in n days, taking 0.85 in sunny days, taking 0.6 in cloudy days and taking 0.4 in rainy and snowy days; The atmosphere mass at time t is Wherein the method comprises the steps of Is the solar altitude.
  3. 3. The method for building the optimal economic model for the capacity allocation upper layer according to claim 1, wherein for the upper layer optimization model, with the minimum total cost in the operation and maintenance period, an objective function consists of two items of the TPSS allocation investment cost and the future operation and maintenance cost, and in the calculation of the allocation investment cost, the capacity of the PV and the capacity and power of the SC are decision variables, and the objective function is as follows: Wherein, the Cost per unit SC capacity Cost per unit power Cost per unit photovoltaic capacity The operation and maintenance cost is; Furthermore, the objective function needs to meet the above constraints.
  4. 4. The method for building an optimal economic model for a capacity configuration lower layer according to claim 1, wherein for the lower layer optimization model, the cost of maintenance is maintained in TPSS period The minimum target is that the key variable is the photovoltaic output power in unit time T Feedback power of power grid Charge-discharge power of SC The minimum maintenance cost is obtained by considering each cost, and the objective function is as follows: Wherein the method comprises the steps of Representing the total cost of maintenance over a maintenance period of N years, For the annual operating cost of the ith year TPSS system, In order to achieve the discount rate, For the inflation rate, D is the actual number of days of operation in the i-th year, For the ith solar photovoltaic operation and maintenance cost, there are For the ith electricity charge Wherein the method comprises the steps of The power is purchased from the power grid in unit time T, and the power is provided with Wherein, the For the load power For photovoltaic output power For the purpose of storing the output power of the power, Is the cost per unit of electrical degree, For the i-th daily electricity charge, there are Wherein: in order to obtain the unit price of the electricity fee, For the average power purchase per minute at any continuous 15 minutes on day i, For the benefit of the ith carbon tax, there is Wherein: representing the income of the unit carbon tax, The energy utilization power of the renewable energy source in the time T, Punishment cost for the ith regenerative braking feedback is that Wherein: Representing unit power penalty costs due to feedback grid power For the total power returned to the grid at time T, For the ith SC charge and discharge maintenance cost, there are Wherein: Respectively representing the maintenance cost of unit power generated by SC charge and discharge and the loss cost of unit service life generated by switching charge and discharge states The charging power is put in for the time T, The charge-discharge switching times of SC in the T moment; Furthermore, the objective function needs to meet the above constraints.
  5. 5. The method for converting an upper and lower bound uncertainty model into a determined model solution by interval linear programming according to claim 1, wherein the working condition is divided according to actual data of the photovoltaic output and a photovoltaic output prediction model, and the parameters are solved by For example, particle swarm algorithm pairs are used Solving the upper and lower bounds, and changing different weather to obtain the solar altitude angle Incident angle of sun The method comprises the steps of obtaining a large amount of photovoltaic sample data under each weather by equal parameters, establishing a proper photovoltaic prediction model, obtaining a large amount of prediction samples to serve a particle swarm iterative algorithm, converting an uncertainty model into a deterministic model by adopting interval linear programming, solving the upper and lower boundaries of the uncertain photovoltaic quantity by calling a linear programming solver CPLEX, and obtaining operation and maintenance cost Upper boundary of And lower cost boundary And transmitting an optimization result obtained based on the photovoltaic prediction model to an upper layer, iterating the upper layer and the lower layer through PSO, and obtaining an optimal light storage configuration scheme when the maximum iteration number or optimal conditions are met.

Description

Traction power supply system photovoltaic uncertainty and energy storage optimal configuration strategy considering regional difference Technical Field The invention relates to the technical field of power systems, in particular to a capacity configuration method of a traction power supply system containing photovoltaic and energy storage. Background Under the background of double carbon, as the installed power generation amount of the photovoltaic new energy source in the electrified railway is continuously improved, not only is a new variable added for the traditional problems of traction load peak impact, high regenerative braking energy, negative sequence current and the like, but also the problems of how to reasonably manage and plan the energy, how to select a proper light storage cooperative control strategy and the like are brought. With the laying of railway power grids, a large number of areas where photovoltaic panels are not installed or are not suitable for installation exist along the railway at present. In order to adapt to complicated and changeable working conditions of the electrified railway, source load fluctuation in different areas is smoothly predicted, and a proper source load fluctuation prediction method is selected according to the actual running working conditions and environment of a load locomotive, so that the method is particularly important in the research of a control strategy. The photovoltaic system and the energy storage system are matched for use, so that unreliability of photovoltaic power generation can be made up, and the energy storage system can absorb excessive photovoltaic energy or provide energy for a power grid through photovoltaic coordination control, so that energy waste is avoided. In the peak clipping and valley filling links, the energy storage system is the most critical ring for bearing energy and conveying energy, and the situation that photovoltaic energy cannot be absorbed due to supersaturation of stored energy or the stored energy is too low to release energy to a power grid can occur probabilistically due to the fact that a control strategy is improper. This "dead zone" phenomenon is often avoided by a suitable control strategy. However, in the photovoltaic energy prediction management of the traction power supply system, the model is generally predicted based on only existing photovoltaic data, and consideration of photovoltaic energy output and capacity allocation based on the photovoltaic energy output is lacking when no photovoltaic data is available. Disclosure of Invention The photovoltaic prediction model which considers the regional difference of various factors such as the geographic environment is established, and the photovoltaic uncertainty consideration is carried out on a large amount of photovoltaic data of the double-layer optimization model, so that the photovoltaic digestion is promoted, and the problems of capacity planning and photovoltaic uncertainty under the condition that the photovoltaic data are unknown are solved. In order to achieve the above object, the present invention adopts the following technical scheme that: Obtaining geographical environment data and actual measurement load data of a traction substation and photovoltaic output data of a photovoltaic power station; building a photovoltaic prediction model based on geographic environment data to predict photovoltaic output; dividing the capacity configuration of the traction power supply system into an upper layer and a lower layer; establishing an upper layer variable model based on a photovoltaic prediction model and existing data for an upper layer; and for the lower layer, converting the upper and lower uncertain models into a determined model solution through interval linear programming. Further, the objective function is: solar photovoltaic output function: The upper layer photovoltaic prediction model aims at simulating photovoltaic output data in a prediction time period, and comprises two factors including regional characteristics and weather characteristics, wherein an objective function is as follows: In the middle of And calculating 1440 when the total photovoltaic output is calculated in one day by taking minutes as a measurement unit according to the solar irradiance function at the t moment. (1) Direct irradiance of headroom Wherein a is a constant, generally 0.14; h represents the altitude of the place where the photovoltaic power station is located; the incident angle of sunlight at time t. The atmospheric mass at the time t is calculated as follows: In the middle of Is the solar altitude at time t. (2) Weather type In the middle ofThe weather index value of the ith day is taken as a value according to the weather in the n days, and the range is 0-1. The composition is taken at 0.85 on sunny days, at 0.6 on cloudy days and at 0.4 on rainy and snowy days. Photovoltaic data prediction model correction: For the existing photovoltaic output data are