Modeling And Control Of Mcrogrid Based On Load Response

Micro grid is a small distribution system applied to the distributed power supply, can be run with the external grid, and can also run in island operation when the fault appear in the grid. Micro grid, with residential loads, is close to the end users who are most for commercial and residential users, and the electric equipment is mostly air conditioning and refrigerators and other home furnishing type thermostatically-controlled load with flexible working mode, this type load is a constant temperature heater controlled load, its driving power almost uses AC asynchronous motor comparing to resistive load, has a larger proportion and need reactive power excitation. So after the load response control, it can be used as inactive and reactive power reserve. Since the problem of network loss of power system is prominent, and reactive power compensation is an effective means in reducing the line loss. Therefore, this paper uses BP neural network to proced reactive power optimization of micro grid, combined with load response, in the study of reactive power optimization method of large power grid based on the MATLBA2013a environment.Using the method of equivalent modeling of traditional micro grid, analysised the P/Q, V/f and droop control strategy of the micro power, and also presented the master-slave, peer-peer and layering control methods of micro grid, studied the different control strategies of the energy storing device under grid connected micro grid and islanded micro grid, means current and V/f control strategy, and built a micro grid simulation platform in the MATLAB/Simulink environment, the simulation results verified the rationality and feasibility of the control strategy, and the voltage and frequency of islanded micro grid could meet the national standards.Taking the refrigerator as an example, this paper proposed a kind of variable participation control strategy suitable of thermostatically-controlled loads, and established its mathematical model, constructed thermostatically-controlled load module, run in the micro grid simulation platform. The simulation results showed that, improving the participation of users was to reduce the user comfort for the price, users could choose appropriate participation coefficient according to the comfort requirements. Comparing the results with and without considering the load response, we could find thermostatically-controlled loads, has a hand in power control, not only can satisfy the user’s enthusiasm, but also effectively reduce the battery charge and discharge depth, capacity requirements, prolong the service life, at the same time of reducing the supply of large power grid burden.Extracteing the terminal voltage, active power and reactive power, active power loss and node voltage of micro power generator inside the micro grid, trained the adaptive neural network after normalization of data processing, and produced the on/off amount of capacitor and micro power voltage. Comparing the results of micro grid power flow before and after optimization, this optimization algorithm could not only control the micro grid voltage in the reasonable scope, but also could reduce the system power loss. On this basis, considering thermostatically-controlled loads which taked the refrigerator as the representative, by changing the user participation coefficient, could get the results:the user participation coefficient was higher, the user participation was higher, the thermostatically-controlled loads as reactive power reserve could provide more reactive power, the losses of system was lower, and the voltage verged to the rated voltage, it verified the rationality and feasibility of the optimization method.