Microgrid control system and static synchronous compensation control device based on petri legendre fuzzy neural network

Abstract

This invention provides a microgrid control method and a static synchronization compensation control method based on a Patrice fuzzy neural network. The static synchronization compensation control method includes the following steps: configuring an energy storage circuit to supply current to multiple output nodes; configuring a compensation control circuit to calculate virtual power based on the voltage and current of the output nodes, and calculating the error value between the virtual power and the commanded virtual power; configuring a Patrice compensation circuit to generate a Patrice compensation value based on the error value and its derivative using the Patrice fuzzy neural network; using the compensation control circuit to output compensation command parameters based on the Patrice compensation value; and configuring a distribution-type static synchronization compensator to compensate the output voltage and output current of each output node according to the compensation command parameters.

Claims

1. A microgrid control method based on a Patrice fuzzy neural network, comprising the following steps: configuring an energy storage circuit to supply multiple output currents to multiple output nodes respectively; configuring a sensing circuit to sense the output current flowing to each of the output nodes and an output voltage of each of the output nodes; configuring a compensation control circuit to calculate an output virtual power based on the sensed output voltages and output currents; using the compensation control circuit to calculate an error value between the output virtual power and a command virtual power as a virtual power error value; configuring a Patrice compensation circuit to generate a Patrice compensation value based on the virtual power error value and a virtual power differential error value through a Patrice fuzzy neural network, wherein the virtual power differential error value is a differential value of the virtual power error value; using the compensation control circuit to output a compensation command parameter based on the Patrice compensation value. Using the energy storage circuit, the output current is modulated according to the compensation command parameters; and using the energy storage circuit, the modulated output currents are supplied to multiple output nodes respectively to compensate the output voltage of each output node.

2. The microgrid control method based on a Patrice-Lejson fuzzy neural network as described in claim 1 further comprises the following steps: configuring an input layer, a membership function layer, a Lejson layer, a Patrice layer, a rule layer, and an output layer in the Patrice-Lejson fuzzy neural network; receiving the virtual power error value and the virtual power differential error value from the compensation control circuit via the input layer and the Lejson layer; and outputting the Patrice-Lejson compensation value to the compensation control circuit via the output layer.

3. The microgrid control method based on a Patrice-Leandro fuzzy neural network as described in claim 2 further includes the following steps performed using the Patrice-Leandro compensation circuit: Inputting the virtual power error value and the virtual power differential error value as two neural network input values ​​to the input layer and the Leandro layer; Transmitting each of the neural network input values ​​from the input layer to each of the plurality of neurons in the membership function layer; In the membership function layer, fuzzifying each of the neural network input values ​​to generate a plurality of membership layer output values ​​and outputting them to a plurality of Patrice positions in the Patrice layer; In the Leandro layer, processing the plurality of neural network input values ​​according to a plurality of Leandro expansion coefficient values ​​to obtain a plurality of Leandro expansion values; In the Leandro layer, multiplying the plurality of Leandro expansion values ​​by a plurality of Leandro weight values ​​to obtain a plurality of Leandro layer output values; In the Patrice layer, filtering the plurality of membership layer output values ​​according to a Patrice threshold value to obtain a plurality of Patrice layer output values. In the rule layer, a merging function is used to merge the output values ​​of each of the Lerang layers with a portion of the output values ​​of the multiple Paicui layers to obtain multiple rule layer output values; and in the output layer, the multiple rule layer output values ​​are processed according to multiple rule weight values ​​to obtain multiple output weight values, and the multiple output weight values ​​are added together to obtain the Paicui Lerang compensation value.

4. The microgrid control method based on a Patrice fuzzy neural network as described in claim 3 further includes the following steps: using the compensation control circuit, setting the Patrice threshold value according to the following equation: , where dip represents the Patrice threshold value, and and are each a preset Patrice coefficient, and H represents an average value of the virtual power error value and the virtual power differential error value.

5. The microgrid control method based on a Patrice fuzzy neural network as described in claim 1 further includes the steps of performing the compensation control circuit: calculating a difference between an angular frequency of each of the output voltages and a rated angular frequency of the energy storage circuit, as a rated angular frequency difference; converting the rated angular frequency difference into an angular frequency droop compensation value based on a droop coefficient; setting a real power command compensation value based on the angular frequency droop compensation value and a rated real power of the energy storage circuit; calculating an output real power based on the sensed output voltages and output currents; compensating the output real power based on the real power command compensation value to generate a droop control compensation value; and modulating the compensation command parameters based on the droop control compensation value.

6. The microgrid control method based on Patrice fuzzy neural network as described in claim 5 further includes the following steps: using the compensation control circuit, adjusting the compensation command parameters according to the current flowing from the multiple output nodes to the multiple loads sensed by the sensing circuit.

7. The microgrid control method based on a Patrice fuzzy neural network as described in claim 5 further includes the steps of performing the compensation control circuit: filtering a plurality of the output voltages to generate a plurality of first-order filtered voltages; modulating the plurality of the first-order filtered voltages based on the angular frequency to output a plurality of second-order filtered voltages respectively; generating a first initial phase-locked voltage based on one of the plurality of second-order filtered voltages and one of the plurality of first-order filtered voltages; converting the first initial phase-locked voltage into a first positive-sequence component voltage based on a first phase-locked coefficient; generating a second initial phase-locked voltage based on another of the plurality of second-order filtered voltages and another of the plurality of first-order filtered voltages; converting the second initial phase-locked voltage into a second positive-sequence component voltage based on a second phase-locked coefficient; and generating the angular frequency based on the first positive-sequence component voltage or the second positive-sequence component voltage.

8. The microgrid control method based on a Patrice fuzzy neural network as described in claim 1, further comprising the steps of: taking three of the plurality of output voltages as three output voltages on a three-phase fixed coordinate axis and converting them into two output voltages on a synchronous rotating coordinate axis; taking three of the plurality of output currents as three output currents on the three-phase fixed coordinate axis and converting them into two output currents on the synchronous rotating coordinate axis; and calculating the output virtual power based on each of the output voltages and each of the output currents on the synchronous rotating coordinate axis.

9. A static synchronization compensation control method based on a Patrice fuzzy neural network, comprising the following steps: configuring an energy storage circuit to supply multiple output currents to multiple output nodes respectively; configuring a sensing circuit to sense the output current flowing to each of the output nodes and an output voltage of each of the output nodes; configuring a compensation control circuit to calculate an output virtual power based on the sensed output voltages and output currents, and calculating an error value between the output virtual power and a command virtual power as a virtual power error value; configuring a Patrice compensation circuit to generate a Patrice compensation value based on the virtual power error value and a virtual power differential error value through a Patrice fuzzy neural network, wherein the virtual power differential error value is a differential value of the virtual power error value; using the compensation control circuit, outputting a compensation command parameter based on the Patrice compensation value; and configuring a distribution-type static synchronization compensator to compensate the output voltage and output current of each of the output nodes according to the compensation command parameter.

10. The static synchronization compensation control method based on the Patrice-Lejon fuzzy neural network as described in claim 9 further comprises the following steps: configuring an input layer, a membership function layer, a Lejon layer, a Patrice layer, a rule layer, and an output layer in the Patrice-Lejon fuzzy neural network; receiving the virtual power error value and the virtual power differential error value from the compensation control circuit via the input layer and the Lejon layer; and outputting the Patrice-Lejon compensation value to the compensation control circuit via the output layer.

11. The static synchronization compensation control method based on a Patrice-Leandro fuzzy neural network as described in claim 10 further comprises the following steps performed using the Patrice-Leandro compensation circuit: inputting the virtual power error value and the virtual power differential error value as two neural network input values ​​to the input layer and the Leandro layer; transmitting each of the neural network input values ​​from the input layer to each of the plurality of neurons in the membership function layer; in the membership function layer, fuzzifying each of the neural network input values ​​to generate a plurality of membership layer output values ​​and outputting them to a plurality of Patrice positions in the Patrice layer; in the Leandro layer, processing the plurality of neural network input values ​​according to a plurality of Leandro expansion coefficient values ​​to obtain a plurality of Leandro expansion values; in the Leandro layer, multiplying the plurality of Leandro expansion values ​​by a plurality of Leandro weight values ​​to obtain a plurality of Leandro layer output values; in the Patrice layer, filtering the plurality of membership layer output values ​​according to a Patrice threshold value to obtain a plurality of Patrice layer output values. In the rule layer, a merging function is used to merge the output values ​​of each of the Lerang layers with a portion of the output values ​​of the multiple Paicui layers to obtain multiple rule layer output values; and in the output layer, the multiple rule layer output values ​​are processed according to multiple rule weight values ​​to obtain multiple output weight values, and the multiple output weight values ​​are added together to obtain the Paicui Lerang compensation value.

12. The static synchronization compensation control method based on a Patrice fuzzy neural network as described in claim 11 further includes the following steps: using the compensation control circuit, setting the Patrice threshold value according to the following equation: , where dip represents the Patrice threshold value, and and are each a preset Patrice coefficient, and H represents an average value of the virtual power error value and the virtual power differential error value.

13. The static synchronization compensation control method based on a Patrice fuzzy neural network as described in claim 9 further comprises the following steps performed using the compensation control circuit: calculating a difference between an angular frequency of each of the output voltages and a rated angular frequency of the energy storage circuit, as a rated angular frequency difference; converting the rated angular frequency difference into an angular frequency droop compensation value based on a droop coefficient; setting a real power command compensation value based on the angular frequency droop compensation value and a rated real power of the distribution-type static synchronization compensator; calculating an output real power based on the sensed output voltages and output currents; compensating the output real power based on the real power command compensation value to generate a droop control compensation value; and modulating the compensation command parameter based on the droop control compensation value.

14. The static synchronization compensation control method based on the Patrice fuzzy neural network as described in claim 13 further includes the following steps: using the compensation control circuit, adjusting the compensation command parameters according to the current flowing from the multiple output nodes to the multiple loads sensed by the sensing circuit.

15. The static synchronization compensation control method based on a Patrice fuzzy neural network as described in claim 13 further comprises the following steps performed using the compensation control circuit: filtering a plurality of the output voltages to generate a plurality of first-order filtered voltages; modulating the plurality of the first-order filtered voltages based on the angular frequency to output a plurality of second-order filtered voltages respectively; generating a first initial phase-locked voltage based on one of the plurality of second-order filtered voltages and one of the plurality of first-order filtered voltages; converting the first initial phase-locked voltage into a first positive-sequence component voltage based on a first phase-locked coefficient; generating a second initial phase-locked voltage based on another of the plurality of second-order filtered voltages and another of the plurality of first-order filtered voltages; converting the second initial phase-locked voltage into a second positive-sequence component voltage based on a second phase-locked coefficient; and generating the angular frequency based on the first positive-sequence component voltage or the second positive-sequence component voltage.

16. The static synchronous compensation control method based on a Patrice fuzzy neural network as described in claim 9 further comprises the following steps performed using the compensation control circuit: taking three of the plurality of output voltages as three output voltages on a three-phase fixed coordinate axis and converting them into two output voltages on a synchronous rotating coordinate axis; taking three of the plurality of output currents as three output currents on the three-phase fixed coordinate axis and converting them into two output currents on the synchronous rotating coordinate axis; and calculating the output virtual power based on each of the output voltages and each of the output currents on the synchronous rotating coordinate axis.

17. The static synchronization compensation control method based on a Patrice fuzzy neural network as described in claim 9 further comprises the following steps: In the distribution-type static synchronization compensator, a plurality of upper bridge switches are configured, each upper bridge switch having a first terminal, a second terminal, and a control terminal, the first terminal of each upper bridge switch being connected to the positive terminal of a voltage source; In the distribution-type static synchronization compensator, a plurality of lower bridge switches are configured, each lower bridge switch having a first terminal, a second terminal, and a control terminal, the second terminal of each lower bridge switch being connected to the negative terminal of the voltage source, the plurality of second terminals of the plurality of upper bridge switches being respectively connected to the plurality of first terminals of the plurality of lower bridge switches, and respectively connected to the plurality of output nodes; and In the distribution-type static synchronization compensator, an energy storage element is configured, wherein the two ends of the energy storage element are respectively connected to the positive terminal and the negative terminal of the voltage source; and The plurality of compensation command parameters are received from the compensation control circuit by the plurality of upper bridge switches and the plurality of lower bridge switches.

18. The static synchronization compensation control method based on the Patrice fuzzy neural network as described in claim 17 further includes the following steps: using the compensation control circuit, based on a plurality of compensation command parameters, outputting a plurality of pulse width modulation signals to a plurality of control terminals of a plurality of upper bridge switches and a plurality of control terminals of a plurality of lower bridge switches respectively.

Images