Reactive Power Optimization of Sub-Blubber Rod in Active Distribution Network Based on Improved Grey Markov Model

A wattless power optimization method for active energy distribution networks based on an improved grey Markov model was studied to improve the voltage drop caused by unreasonable wattless power distribution in centralized electricity supply modes. Select the normalized residual sequence method to modify the non-biased grey GM(1,1) model, use the non-biased grey GM(1,1) model that has completed the residual correction to obtain the system voltage index sequence, and use the Markov model to divide the system voltage index sequence Use the Markov state to establish an improved grey Markov model to obtain the prediction results of the system voltage of centralized electricity supply mode; take the lowest deviation of the system voltage prediction control as the goal, and establish the goal of the wattless power output of centralized electricity supply mode Function to solve the objective function under the condition of satisfying the system voltage constraint and the wattless power adjustment capacity constraint of centralized electricity supply mode, and realize the wattless power optimization of centralized electricity supply mode. The case analysis results indicate that this research method can achieve wattless power optimization of distribution networks.


Introduction
Nowadays, new energy power generation has become an important development trend.Local options improve the use of energy in the category of functions and reduce the use of energy.[1][2][3], but it cannot guarantee the optimal operating energy quality.When the power output of distributed power sources fluctuates significantly, it will cause voltage mutations and power flow exceeding limits [4], resulting in frequent short-term voltage exceeding limits.The voltage control problem has attracted the attention of many researchers, such as Zhang et al. [5].Gao et al. [6] proposed a distributed distribution network wattless power optimization method using an improved NSGA-II algorithm.This method can efficiently solve the problem and has been experimentally verified to be practical.A distributed distribution network response optimization method based on the improved NSGA-II algorithm [6] was proposed, with a focus on studying the impact of distributed capacity uncertainty on the distribution network.The above two methods are superior to the optimization process, which is too complex and computationally expensive.Their optimization effect is not ideal, and the cost is on the high side.Based on this, in order to improve the operational stability of the active energy distribution network, the distributed robust wattless power optimization method of the active distribution network based on the improved grey Markov model is studied.We select an improved grey Markov model to predict the system voltage and use the system voltage prediction results to achieve robust wattless power optimization of centralized electricity supply mode distribution, avoiding frequent fluctuations in the system voltage of centralized electricity supply mode and ensuring that the system voltage is stable near the reference voltage.As the use of the queue method is applied to centralized electricity supply mode examples, it is verified that the search method is most effective in optimizing the active distribution power of the distribution network and can ensure reliable operation.Power grid distribution system.

An improved bus-line voltage prediction of the grey Markov model 2.1.1. Unbiased gray GM (1,1) model
We establish the traditional grey system GM (1,1) model of the raw data sequence of the active distribution network.
( ) ( ) In the formula, A and Yn are centralized electricity supply mode system voltage data matrices, λ and K are the initial sequence parameters of centralized electricity supply mode system voltage, and a and u are both parameters of the traditional grey system GM(1,1) model.

The active sales network
We establish a non-biased grey GM (1,1) model using the initial sequence of system voltage in an active energy distribution network as follows: (2) The process of establishing a non-biased grey GM (1,1) model for the initial sequence of system voltage is as follows.
(1) We establish unbiased expressions for the traditional grey system GM (1,1) model; (2) We find the parameters of the non-biased grey GM (1,1) model: (3) We establish an exponential sequence model of the raw data: The non-biased grey GM (1,1) model eliminates the need for incremental reduction operations, shortens the modelling steps, and improves the computational efficiency of the model.

Correct the residue by the normalized residue sequence method
In order to improve the applicability of the non-biased grey GM (1,1) model in robust wattless power optimization of active energy distribution network, its residual data sequence was modified.The specific process is as follows: (1) We obtain the deviation sequence of the model.
We obtain the normalized residue sequence of the model.
The final non-biased grey GM (1,1) model is determined as follows:

Improve the grey Markov prediction model
The improved grey prediction model is used to achieve an accurate prediction of system voltage in an active energy distribution network.The model establishment process is as follows.
(3) We establish a state transition probability matrix P , where the expression for the element in the matrix P is as follows: In Formula (9), ( ) We set the corresponding state of the object as the initial state according to the order of the distance from small to large and search the corresponding row vector of each object in the transition probability matrix.We establish a new probability matrix expression as follows.
We establish an improved grey Markov system voltage prediction model expression as follows:

Realizing unbiased robust wattless power optimization
Rolling optimization of wattless power output in active sales networks using system voltage prediction results, the wattless power optimization process is as follows.
(1) We establish the objective function By selecting the system voltage as the basis in time and selecting the minimum deviation of system voltage predictive control as the objective function for robust wattless power optimization in active sales network allocation, an objective function for robust wattless power optimization in active sales network allocation can be established.The specific form is as follows: In Formula (12), represent the predicted vector of system voltage and the reference vector of system voltage at time t , respectively.
(2) Constraint conditions (a) System voltage constraint When adjusting the system voltage in an active sales network, it is necessary to simultaneously adjust the system voltage of the power collection and the distributed distribution stations connected to the power system.The system voltage constraints are set as follows: In Formula (13), 1   U and pv U represent the system voltage control values collected by the centralized electricity supply mode system and the system voltage control values of the distributed distribution station, respectively.(b) Constraints on wattless power regulation capability of active sales networks as follows: In Formula ( 14), ,max svc Q and ,min svc Q represent the upper and lower limits of wattless power output in centralized electricity supply mode, respectively.

Example analysis
We select the centralized electricity supply mode of a certain power enterprise as the research object and simulate the operation of the power system using the Matlab simulation platform.

Performance test
This article adopts an improved grey Markov prediction model to predict the system voltage of active sales networks.The statistical method used in this article predicts the system voltage of centralized electricity supply mode operating for 35 minutes, as well as the system voltage of distributed power sources.The system voltage prediction results are shown in Figure 1. the actual voltage operation results of the centralized electricity supply mode, it can be seen that the system voltage prediction results of this method are very close to the actual system voltage operation results.The method proposed in this article has extremely high performance in predicting system voltage, utilizing high prediction results to ensure robust wattless power optimization of active sales network distribution.The method proposed in this article not only effectively predicts the system voltage collected by the system, but also has a high predictive effect on the system voltage of distributed power sources in active sales networks, providing a basis for wattless power optimization in active sales networks.

Comparative testing
Next, the methods of [5] and [6] are selected as comparative methods to conduct comparative tests.
After applying the three methods, the wattless power loss and measured node voltage deviation of the centralized electricity supply mode are statistically analyzed.The statistical results are shown in Table 1.From Table 1, it can be seen that, through the method of this article, the wattless power output of the centralized electricity supply mode is optimized considering the system voltage.The wattless power loss of different measurement nodes is lower than 20 kw, and the node voltage deviation is lower than 0.02 p.u.However, using two comparison methods, the different measurement values are lower than 0.03 p.u.Therefore, the wattless power loss and voltage deviation of the wattless power output of centralized electricity supply mode optimized using the method proposed in this paper are significantly lower than those of the application comparison method, verifying that the robust wattless power optimization method for active sales network distribution proposed in this paper has extremely high optimization effectiveness.

Conclusions
This study proposes an improved, robust wattless power optimization method for solving wattless power optimization problems in active sales networks.This method uses an improved grey Markov prediction model to predict system voltage and reduce control bias through the results, thereby avoiding frequent fluctuations in system voltage.This method can accurately predict system voltage and achieve robust power sales by optimizing the allocation of wattless power.The experimental results indicate that this method can effectively reduce system voltage fluctuations and achieve wattless power optimization in active power distribution networks.

Figure 1 .
Figure 1.System Voltage Prediction ResultsFrom the experimental results in Figure1, it can be seen that the improved grey Markov prediction model used in this paper can effectively predict system voltage.Comparing the predicted results with

Table 1 .
Comparison of wattless power Optimization Effects