Font Size: a A A

Research On Key Control Strategy Of Medium-low Voltage DC Distribution Network

Posted on:2022-09-24Degree:DoctorType:Dissertation
Country:ChinaCandidate:X G HuangFull Text:PDF
GTID:1482306560993609Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
Compared with the traditional AC power grid,DC power grid has the advantages of convenient renewable energy access,high power supply efficiency,large power supply capacity and long supply radius,small power supply corridor area and excellent power quality.It has been verified and widely used in the field of HVDC transmission and LVDC distribution.In recent years,with the increasing penetration of renewable energy power generation system and the rapid growth of urban load,the demand for Medium Voltage DC distribution network(MV DCDN)construction is gradually increasing.MV DCDN has many voltage levels,complex network structure and various operation modes.Reasonable control strategy is the key to ensure its safe and stable operation.This paper focuses on the typical MV DCDN with distributed energy resource(DER)access,and studies key technologies that affect its stability and power flow optimization,such as coordinated control,droop control,virtual inertia support,and optimal dispatch.The details are as follows:Aiming at the coordination control problem of DC distribution network interconnected by DC solid state transformer(DCSST)with different voltage levels,a hybrid coordinated control method based on power margin and voltage deviation slope is proposed.The information of voltage deviation and power margin is used for pattern recognition of power grid,which effectively alleviates the contradiction between the system with too small voltage deviation easy to lose stability and the system with too large voltage deviation difficult to expand.The introduction of voltage slope transition control can effectively suppress the DC voltage overshoot during the system mode switching,improve the stability and shorten switching time.Based on the average equivalent model method and the hybrid EMF large signal modeling method,the model of DCSST and voltage source converter(VSC)are established.Then,a control strategy stability analysis method based on the hybrid EMF model of DC distribution system is proposed to guide the design of coordination controller parameters of key equipment such as the above.To solve the problems of power distribution accuracy of grid connected nodes with droop control in DCDN,a dynamic consistent droop control strategy based on power predictive control is proposed in this paper,which can realize high-precision and dynamic power distribution of grid connected VSC.At the same time,the overall control loop is simplified compared with the traditional adaptive droop method,which reduces the complexity of control.This method introduces power predictive control into the VSC inner control,which can improve the dynamic response and tracking speed when the power or load fluctuates;on the other hand,the secondary coordinated control can realize power balance by taking the power as the interaction information between devices.The secondary coordinated control combines the traditional droop control with dynamic consensus algorithm(DCA),introduces the initial state variables and weight coefficients of the system in DCA,which makes the judgment of convergence conditions easier.The small signal model of the system with power prediction and dynamic consistency algorithm is derived.The root locus method is used to analyze the influence of the control strategy on the line impedance,communication delay and other parameters on the system stability,so as to guide the design of distributed droop system controller.Focus on the weak inertia problem of DC distribution network,this paper proposes an improved virtual capacitor(IVC)method in equipment virtual inertia support strategy,which uses the information of power and voltage change rate to adjust the virtual capacitor in real time,so as to realize the flexible adjustment of equipment virtual inertia,and improve the voltage quality and anti-disturbance ability of DC distribution network.At the same time,aiming at the problem that how multiple nodes having inertial support ability cooperate with each other,the output impedance of the equipment and the virtual inertia of the system are adjusted by interactive power,virtual impedance and other converters' state information by low speed communication,so as to improve the dynamic and steady-state control characteristics of the system.Based on the average equivalent model method,the small signal model of DC power grid based on IVC control is constructed.Through root locus analysis,the influence of main control parameters on system virtual inertia and DC voltage stability is revealed to guide the selection of equipment controller parameters.Aiming at the energy multi-objective optimization problem of DCDN with multi DERs access,a multi-mode and multi-time scale multi-objective optimal dispatching control strategy for DCDN is proposed to realize the comprehensive optimization objectives which include maximizing the renewable energy consumption,minimizing the network loss and minimizing the load loss in DC power grid.The day-ahead and intra-day dispatching models are established respectively;the optimization weight coefficient is dynamically adjusted according to the operation mode,so that the day-ahead multi-objective optimization problem is transformed to a single objective optimization problem,and finally solved with quadratic constrained quadratic programming method.Model predictive control(MPC)is used to solve the correction of the day-ahead scheduling instructions,and the improved Particle Swarm Optimization(PSO)algorithm is adopted to solve intra-day optimal problem.By introducing compression factor and adaptive inertia coefficient,the global optimization ability of PSO algorithm is enhanced,so as to ensure that the whole scheduling strategy responds to changes in the environment and load in time and improves the effect of power flow optimization.
Keywords/Search Tags:medium voltage DC distribution network, low voltage DC distribution network, hybrid coordinated control, droop control, dynamic consistency algorithm, model predictive control, improved virtual capacitor, multi-objective optimization
PDF Full Text Request
Related items