Research on improved linear auto-disturbance rejection control of hybrid energy storage in optical storage microgrid

Photovoltaic power generation has uncertainties such as randomness and volatility. In order to ensure the stable operation of the power system, hybrid energy storage technology is introduced to improve the voltage stability of the optical storage microgrid. Traditional PI control still needs some help in response speed and overshoot. Facing the hybrid energy storage microgrid system with nonlinear and strong coupling characteristics, in order to improve power quality and reduce bus voltage fluctuation and impact, this paper designs an improved active disturbance rejection control (ADRC) integrated energy storage control system based on a fuzzy algorithm, Based on the traditional double closed-loop PI control, the fuzzy LADRC is used to replace the voltage outer loop, and the system model is built on the MATLAB/Simulink platform for simulation. The results show that the proposed control strategy has a smaller overshoot, faster dynamic response time, and good anti-interference ability compared with the traditional PI and ADRC, which improves the dynamic performance of the photovoltaic energy storage system while stabilizing the bus voltage fluctuation and ensures the stable operation of the powergrid.


Introduction
With the depletion of traditional fossil energy and its adverse impact on the environment, changes in the energy structure have followed.Wind power and photovoltaic power generation play a leading role in new energy, but their uncontrollable randomness and volatility challenge the stable operation of the power system [1] .The rapid development of microgrid technology plays an indispensable role in efficiently using renewable energy [2] .In order to solve the problem of DC bus voltage fluctuation caused by the unstable output power of photovoltaic power generation system, it is necessary to configure a specific capacity of energy storage system to suppress the power fluctuation of microgrid and maintain the bus voltage balance.
In recent years, applying auto-disturbance rejection control (ADRC) as a control algorithm in DC-DC converters has aroused extensive research.Active disturbance rejection control (ADRC) inherits the essence of PID control with a simple structure, strong robustness, and strong anti-interference ability.Document [3] proposes a control strategy for the energy storage converter of the superconducting magnetic energy storage system based on linear auto-disturbance rejection control.It uses linear ADRC to compensate for the system disturbance, improve the output power quality of the energy storage system, and improve the system robustness.The root locus method is used to analyze its stability and robustness [4] [5] .
Based on the above problems, this paper improves the linear ADRC, and uses it to replace the voltage outer loop control of the dual closed-loop PI control, introduces fuzzy control, and carries out online self-tuning adjustment of the control parameters of the ADRC.Finally, through Simulink simulation and comparison [6] .
2. Basic structure of active disturbance rejection control As an upgraded control strategy of PID control, ADRC inherits its essence.By introducing the state observation technology, a state observer is constructed using the input and output of the system to estimate the total disturbance in real time and compensate in the feedback control link [7] .Active disturbance rejection control mainly consists of tracking differentiator (TD), linear state feedback (LSEF), and extended state observer (ESO).The overall structure diagram of LADRC is shown in Figure 1.

3.1.Tracking differentiator
The function of the tracking differentiator is to process the error signal, and its purpose is to arrange an appropriate transition process for the input signal to facilitate real-time tracking.
When the DC bus voltage is 400V, the system is simulated to obtain the output waveform of the tracking differentiator, as shown in Figure 2 and Figure 3.
As shown in Figure 2, when the velocity factor r is constant, the smaller the h is, the weaker the filtering ability is, and the larger the h is, the slower the response speed of  (t) will be.As shown in Figure 3, when the value of filtering factor h is constant, the smaller r is, the faster the response speed is, but the larger r value will cause a larger peak value of  (t).So,in this system, h is 0.1, and the r is 10.

3.2.Design of extended state observer
The differential equations of common second-order systems can generally be: as system disturbance.
The corresponding third-order LESO is: The observer bandwidth is set as ω , configure the upper pole to ω , that is: It can be calculated that: β = 3ω , β = 3ω , β = ω

3.3.Improved design of linear state error feedback control rate
The classical PID control is as follows: This paper's improved linear state error feedback control rate is obtained from the traditional PID control idea.The linear gain parameters kp and kd are regarded as the gains of the proportional and differential links in PID control.The error value e1 is obtained by the difference between the output v1 generated by the TD link and the output z1 generated by the extended observer (ESO) link, and e0 is obtained by integrating it [8] .It is equivalent to introducing the gain parameter ki of the integration link.Compared with the traditional control rate, an integration link is added after the improvement [9] .

3.4.Add fuzzy control to improved LADRC
In this paper, fuzzy adaptive auto-disturbance rejection control is adopted.Referring to the fuzzy PI control, the fuzzy controller inputs the input error variable and the differential error change to the auto disturbance rejection controller after fuzzy adjustment [10] .The output is the parameter increase , of the proportional differential controller [11] .
It can be concluded that: Where The schematic diagram of fuzzy adaptive auto-disturbance rejection control for a hybrid energy storage system is shown in Figure 5.

Simulation Analysis
This paper simulates and analyzes the microgrid system under three kinds of control.The specific parameters of the energy storage system are shown in Table 1, and the controller parameters of LADRC are as follows.
In the transition process TD, the filter factor h is set to 0.1, and the velocity factor r is set to 10.In extended observer ESO, observer bandwidth ω =1000rad/s controller bandwidth in linear error feedback LSEFω =1000rad/s,b0=20000.
The parameters of the battery current inner loop PI controller are Kp1=0.2,Ki1=25,and the super capacitor current inner loop PI controller are Kp2=1.5,Ki2=1000.Condition1: PV disturbance simulation analysis Before0.2s,Pload=3kW, light intensity S=600W/m 2 ,the illumination intensity increased sharply at 0.2s to 1200W/m 2 , the photovoltaic output increased, and the bus voltage recovered to the reference value after a temporary increase.As shown in Figure 6, after the comparison of three control strategies, the response speed of the improved control strategy proposed in this paper is faster from startup to the first time that the DC bus voltage is stable, which is more than twice as fast as the traditional control speed, and only takes about 0.015s to reach the stable state.The overshoot is also reduced to a certain extent, and the maximum fluctuation value is reduced to about 10V.When the light intensity suddenly increases, the photovoltaic output exceeds the load demand, the DC-DC converter is in the Buck mode, and the energy storage system is charged.Under the traditional PI control, the maximum voltage fluctuation is 20V, and the traditional LADRC control is 10V.However, the improved LADRC control reduces the maximum fluctuation to about 7V and reduces the transient time by 0.013s.
Condition 2: load disturbance simulation analysis After 0.2s, the light intensity S=1200W/m 2 remains unchanged.When the system runs stably to 0.5s, the DC load Pload suddenly increases from 3kW to 4kW, and when it reaches 0.7s, the DC load decreases to 2.5kW.

Figure7. Bus voltage under photovoltaicfluctuation
As shown in Figure 7, when the load suddenly increases in 0.5s, the peak value of DC bus voltage fluctuation under the traditional PI control is about 10V, and the time to recover to the steady state is about 0.1s.The peak value of DC bus voltage fluctuation under the traditional LADRC control is not significantly reduced.However, the time to recover to the steady state of the system is reduced to 0.02s.Under the improved fuzzy LADRC control, the system is less affected by the load change, the peak value is only about 5V, and the time is shorter to less than 0.01s.Similarly, in 0.7s, when the load decreases, the excess energy is stored in the energy storage equipment, but the effect of the traditional PI control is not ideal.The peak value reaches 15V, and after 0.1s, the DC bus voltage is adjusted as the reference value.The amplitude and time of the traditional LADRC are reduced.However, the improved control strategy proposed in this paper makes the instantaneous value of the DC bus increase only 4V, and the system returns to normal after 0.01s.
The above two conditions can show that the improved LADRC control strategy can achieve good results in reducing voltage fluctuations.

Conclusion
This paper briefly introduces the characteristics of a DC microgrid.It proposes an improved LADRC control strategy based on fuzzy theory to solve the problem of voltage fluctuation and the impact of DC buses.The linear auto-disturbance-rejection voltage outer loop controller is used to replace the original PI control, and fuzzy inference is made.Finally, the control strategy is compared with other control strategies and the following conclusions are drawn: (1) Under the same light intensity and other conditions, the scheme proposed in this paper is superior to other schemes in response time and dynamic response speed.The impact is smaller, and the maximum fluctuation peak is lower.
(2) Compared with the traditional PI control and the traditional LADRC control, the improved LADRC control improves the system stability, the control effect is good, and the robustness is good.In terms of DC bus voltage fluctuation and fluctuating peak current, it has been optimized and reduced to some extent.Its anti-interference ability has been improved, and the system's adaptability has been strengthened.It has an excellent effect in suppressing the voltage fluctuation of the DC bus.

Figure 3 .
Figure2.Tracking the output waveform of a differentiator when h changes

Figure5.
Figure5.Schematic diagram of fuzzy auto disturbance rejection control

Table 1 .
Specific parameters of energy storage system