A zero-carbon building light storage direct-flexible distributed energy system dynamic scheduling method

Abstract

The application discloses a zero-carbon building light storage direct-flexible distributed energy system dynamic scheduling method, and belongs to the technical field of distributed energy scheduling. The method responds to a direct-current distribution bus voltage fluctuation event, collects a photovoltaic maximum power point tracking state, an energy storage power balance state and a flexible load power demand elasticity coefficient through an energy router; generates an initial scheduling instruction according to embedding of a dynamic virtual impedance adjustment function of the photovoltaic and the energy storage state in a direct-current bus voltage droop control loop; inputs the initial scheduling instruction into a mode switching arbitrator, selects an energy scheduling main mode from a photovoltaic priority mode, an energy storage priority mode, a load following mode and a power grid interaction mode in combination with the elasticity coefficient; recalculates function parameters of the dynamic virtual impedance adjustment function according to the main mode, corrects the initial scheduling instruction to obtain a pulse width modulation signal; and distributes the pulse width modulation signal to corresponding bidirectional direct-current converters of the photovoltaic, the energy storage and the flexible load, so that dynamic scheduling is realized.

Description

Technical Field

[0001] This invention relates to the field of distributed energy dispatching technology, specifically a dynamic dispatching method for a zero-carbon building photovoltaic-storage-DC-flexible distributed energy system. Background Technology

[0002] Zero-carbon buildings commonly adopt a photovoltaic-storage-DC-flexible power distribution architecture. Photovoltaic power generation units, energy storage batteries, and flexible loads are coupled via a DC bus, with an energy router handling system power balancing and scheduling. Fluctuations in the DC bus voltage directly reflect power supply and demand imbalances within the system. Existing scheduling methods often rely on droop control with fixed parameters, allocating photovoltaic and energy storage outputs based on voltage deviations using a fixed virtual impedance. Photovoltaic power generation units are affected by environmental factors, and their operating points may fluctuate near their maximum power point or deviate due to power curtailment. Fixed droop coefficients cannot reflect the real-time available power capacity of photovoltaics, leading to curtailment or bus voltage instability. Energy storage units consist of multiple battery cells connected in series or parallel, and differences in the state of charge (SOC) of each cell are common. Fixed droop control treats all cells equally, easily exacerbating power imbalances and shortening the energy storage cycle life. In flexible load management, existing methods often employ on / off control or simple priority unloading, failing to incorporate the demand elasticity represented by the allowable power outage duration and adjustable power range, thus failing to balance building energy comfort with system regulation capabilities. In mode switching strategies, photovoltaic priority, energy storage priority, or grid interaction modes are typically switched rigidly based on bus voltage thresholds, without incorporating flexible load elasticity coefficients to determine the photovoltaic-energy storage-flexible co-operation mode. How to utilize the photovoltaic maximum power point tracking state and the energy storage power balance state to drive the adaptive adjustment of virtual impedance in droop control, and how to introduce flexible load power demand elasticity coefficients in mode switching arbitration to guide the adaptive selection of the main dispatch mode, are problems that remain to be solved. Summary of the Invention

[0003] This paper presents a dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-flexible distributed energy system. It addresses the problems of insufficient droop control adaptability caused by the failure of existing technologies to consider the real-time availability of photovoltaics and the balance of energy storage, as well as the scheduling rigidity caused by the failure to utilize the power demand elasticity of flexible loads during mode switching. The method achieves adaptive energy scheduling based on dynamic virtual impedance and elastic coordination.

[0004] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a dynamic scheduling method for a zero-carbon building photovoltaic-storage-DC-flexible distributed energy system. In response to voltage fluctuation events on the DC distribution bus within a zero-carbon building, the power adaptive interface of the energy router collects the maximum power point tracking (MPPT) status of the photovoltaic power generation units, the energy balance status of the energy storage units, and the power demand elasticity coefficient of the flexible load units in the photovoltaic-storage-DC-flexible network. Based on the MPPT status and energy balance status, a dynamic virtual impedance adjustment function is embedded in the DC bus voltage droop control loop to generate an initial scheduling command for balancing power fluctuations in the photovoltaic-storage-DC-flexible network. This initial scheduling command... The mode switching arbitrator input to the energy router, combined with the power demand elasticity coefficient of the flexible load unit, selects the current energy dispatching master mode from a variety of preset distributed energy collaborative working modes. Based on the selected energy dispatching master mode, the function parameters of the dynamic virtual impedance adjustment function in the DC bus voltage droop control loop are recalculated, and the initial dispatching command is corrected using the obtained function parameters to obtain the pulse width modulation signal used to drive the bidirectional DC converter. The pulse width modulation signal is distributed to the corresponding bidirectional DC converters of the photovoltaic power generation unit, energy storage unit, and flexible load unit, completing the dynamic dispatching of the zero-carbon building photovoltaic-storage-DC-flexible distributed energy system. This method utilizes the real-time status of multiple sources and the elastic collaboration of flexible loads, adaptively adjusts the droop characteristics through dynamic virtual impedance, and switches energy collaboration modes in real time. This can quickly smooth DC bus power fluctuations, improve photovoltaic utilization, maintain the power balance of energy storage units, and fully release the adjustment potential of flexible loads.

[0005] Preferably, in the power adaptive interface of the energy router, real-time data links are established with the communication ports of the photovoltaic DC-DC converter, the energy storage battery management system, and the flexible load controller, respectively. The output voltage and current of the photovoltaic power generation unit are read from the communication port of the photovoltaic DC-DC converter. By comparing the power change direction at adjacent sampling times, it is determined whether the photovoltaic power generation unit is currently in maximum power point tracking (MPPT) mode. The state of charge (SPC) of each energy storage battery cell is read from the communication port of the energy storage battery management system. The difference between the maximum and minimum SPC values ​​of all energy storage battery cells within the energy storage unit is calculated. Based on the comparison result of this difference with a preset balancing threshold, the power balance state of the energy storage unit is determined. The allowable power outage duration and adjustable power range of each flexible load are read from the communication port of the flexible load controller. The product of the allowable power outage duration and the adjustable power range is used as the power demand elasticity coefficient of the flexible load unit. By establishing multiple communication links to synchronously collect multi-dimensional state characteristics, scheduling commands can accurately reflect the real-time operating conditions of the system.

[0006] Furthermore, when generating the initial dispatch command, if the photovoltaic (PV) power generation unit is in maximum power point tracking (MPPT) mode, its output power is taken as the available PV power; if it is not in MPPT mode, its rated power is taken as the available PV power. Based on the energy storage unit's power balance status, when the difference between the maximum and minimum state of charge (SOC) of a single energy storage cell is less than a preset balance threshold, the total discharge capacity of the energy storage unit is recorded as the available energy storage power. When the difference is greater than or equal to the preset balance threshold, the discharge power of the energy storage cell with the lowest SOC is taken as the available energy storage power. The sum of the available PV power and the available energy storage power is used as the reference power for the DC bus voltage droop control loop. This reference power is mapped to a reference voltage offset of the DC bus voltage using a dynamic virtual impedance adjustment function. This reference voltage offset is used to correct the voltage setpoint of the droop control loop, generating the initial dispatch command to balance power fluctuations in the PV-storage-DC-flexible network. This configuration allows for automatic adjustment of the bus voltage reference value when PV power is abundant or energy storage discharge capacity is limited, guiding the rational flow of power.

[0007] In a preferred embodiment of the present invention, the mode switching arbitrator of the energy router pre-sets a photovoltaic priority mode, an energy storage priority mode, a load following mode, and a grid interaction mode as distributed energy collaborative working modes. The power flow direction identifier in the initial scheduling command is parsed as a power injection identifier or a power absorption identifier. When the power flow direction identifier is a power injection identifier, a load reduction factor is calculated based on the product of the reciprocal of the power demand elasticity coefficient and the available photovoltaic power. If the load reduction factor is greater than a first mode threshold, the photovoltaic priority mode is selected as the primary energy scheduling mode for the current moment; otherwise, the energy storage priority mode is selected. When the power flow direction identifier is a power absorption identifier, an energy storage rechargeable factor is calculated based on the product of the power demand elasticity coefficient and the energy storage release capacity. If the energy storage rechargeable factor is greater than a second mode threshold, the load following mode is selected as the primary energy scheduling mode for the current moment; otherwise, the grid interaction mode is selected. This method combines flexible load adjustment capability and adjustable energy storage capacity for dynamic decision-making, adaptively switching to the optimal collaborative mode.

[0008] When recalculating the function parameters of the dynamic virtual impedance adjustment function according to the selected energy dispatch main mode, if the main mode is photovoltaic priority mode, the virtual impedance value is set to the first impedance value, and the voltage offset coefficient is set to the ratio of the available photovoltaic power to the rated voltage of the DC bus; if the main mode is energy storage priority mode, the virtual impedance value is set to the second impedance value, which is greater than the first impedance value, and the voltage offset coefficient is set to the ratio of the energy storage release power to the rated voltage of the DC bus; if the main mode is load following mode, the virtual impedance value is set to the third impedance value, which is between the first and second impedance values, and the voltage offset coefficient is set to the product of the power demand elasticity coefficient and the current voltage of the DC bus; if the main mode is grid interaction mode, the virtual impedance value is set to the fourth impedance value, which is equal to the equivalent impedance of the energy router grid connection interface, and the voltage offset coefficient is set to zero; the reference voltage offset is recalculated using the set dynamic virtual impedance adjustment function, and the recalculated reference voltage offset replaces the corresponding offset in the initial dispatch command to generate a pulse width modulation signal. The mode-adaptive virtual impedance and offset coefficient adjustment match the droop characteristics with available power and load resilience under different operating conditions, thereby enhancing the stability of the DC bus voltage.

[0009] During the distribution of pulse width modulation (PWM) signals, a first drive channel is established in the PWM signal distributor of the energy router, corresponding to the photovoltaic bidirectional DC-DC converter of the photovoltaic power generation unit, a second drive channel corresponding to the energy storage bidirectional DC-DC converter of the energy storage unit, and a third drive channel corresponding to the load bidirectional DC-DC converter of the flexible load unit. When the power flow direction identifier in the PWM signal is a power injection identifier, the duty cycle parameter in the PWM signal is simultaneously written into the first and second drive channels, and the output of the third drive channel is disabled, allowing photovoltaic and energy storage to jointly inject power into the bus. When the power flow direction identifier is a power absorption identifier and the energy dispatch main mode is load following mode, the duty cycle parameter is written into the third drive channel, and the output of the first and second drive channels is disabled, utilizing the flexible load to absorb excess power. When the power flow direction identifier is a power absorption identifier and the energy dispatch main mode is grid interaction mode, the duty cycle parameter is simultaneously written into the first, second, and third drive channels, realizing energy interaction with the grid. By allocating the duty cycle by channel according to mode, coordinated operation of different converters under their respective operating conditions is ensured, channel conflicts are prevented, and power paths are optimized.

[0010] When determining the maximum power point tracking (MPPT) status of a photovoltaic (PV) power generation unit, the output voltage and current values ​​at multiple sampling moments are read from the communication port of the PV DC converter at fixed sampling intervals. The product of the output voltage and current values ​​at each sampling moment is taken as the instantaneous PV power value at that sampling moment. The difference between the instantaneous PV power values ​​at adjacent sampling moments is calculated. If the difference is positive, the power change direction is recorded as upward; if it is negative, it is recorded as downward. The power change direction within three consecutive sampling intervals is counted. If the power change direction is upward or downward for three consecutive intervals and the power change amplitude exceeds a preset disturbance amplitude threshold, the PV power generation unit is determined not to be in MPPT status; otherwise, it is determined to be in MPPT status. This determination method can quickly identify PV operating point offsets caused by changes in the external environment, improving the timeliness of dispatch commands.

[0011] When determining the energy balance state of an energy storage unit, the state of charge (SOC) values ​​of each individual battery cell are read from the communication port of the energy storage battery management system. The maximum and minimum SOC values ​​are extracted, and the difference is calculated. This difference is compared with a preset balance threshold. If the difference is less than the preset balance threshold, the energy storage unit is marked as balanced; if the difference is greater than or equal to the preset balance threshold, it is marked as unbalanced. This accurately reflects the inconsistency of the battery pack and provides a reliable criterion for balance state determination in scheduling.

[0012] When calculating the power demand elasticity coefficient of a flexible load unit, the allowable power outage duration and adjustable power range of each flexible load are read from the communication port of the flexible load controller. The difference between the maximum and minimum adjustable power values ​​within the adjustable power range is taken as the power regulation bandwidth of that load. The product of the allowable power outage duration and the power regulation bandwidth is taken as the power demand elasticity coefficient of that load. The power demand elasticity coefficients of all flexible loads are summed to obtain the power demand elasticity coefficient of the flexible load unit. This coefficient integrates the load's regulation capacity and time constraints, more accurately quantifying the potential of flexible loads to participate in scheduling and regulation.

[0013] In the specific implementation of generating the initial dispatch command, the available photovoltaic power and the dissipable energy storage power are fed into an adder to obtain the reference power. The reference power is then input into a dynamic virtual impedance adjustment function. The virtual impedance value within the function is multiplied by the reference power to form a first intermediate quantity, and the voltage offset coefficient is multiplied by the reference power to form a second intermediate quantity. The two are added together to obtain the reference voltage offset. The base voltage setpoint is read from the voltage setpoint memory of the DC bus voltage droop control loop. The reference voltage offset is subtracted from the base voltage setpoint to obtain the corrected voltage setpoint, which is written into the voltage setpoint register of the droop control loop. This triggers the droop control loop to calculate the initial duty cycle of each bidirectional DC converter and package it into the initial dispatch command. This mechanism dynamically shifts the droop curve based on the system's power generation and energy storage status, optimizing power allocation while maintaining bus voltage safety.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The system collects the output voltage and current of the photovoltaic (PV) power generation units, compares the power change direction at adjacent sampling times to determine the maximum power point tracking (MPPT) status. When the PV is not in MPPT status, the rated power is used as the available power; when it is in MPPT status, the actual output power is used as the available power. Simultaneously, the system obtains the state of charge (SOC) of each cell from the energy storage battery management system, calculates the difference between the maximum and minimum SOC, and uses the total discharge capacity as the energy storage's dischargeable power when the difference is less than the balancing threshold. When the difference exceeds the threshold, the discharge power of the cell with the lowest SOC is used as the energy storage's dischargeable power. The sum of the PV's available power and the energy storage's dischargeable power is used as a reference power, input to a dynamic virtual impedance adjustment function embedded in the droop control loop. The reference voltage offset is calculated by multiplying the virtual impedance value by the voltage offset coefficient and then superimposing the results. This offset is used to correct the droop control voltage setpoint and generate the initial scheduling command. This allows the droop characteristics to dynamically change in real time with the actual available PV capacity and energy storage balancing constraints, eliminating the additional voltage drop caused by fixed impedance parameters when PV power is limited, avoiding over-discharge of weak cells in unbalanced states, and improving the accuracy of PV-energy storage power allocation and the rigidity of the DC bus voltage. In the mode switching arbiter of the energy router, a power demand elasticity coefficient for flexible load units is introduced. This coefficient is obtained by summing the product of the allowable power outage duration and the adjustable power bandwidth of each flexible load. The power flow direction identifier in the initial dispatch command is analyzed. When it is a power injection identifier, the load reduction factor is calculated by multiplying the reciprocal of the elasticity coefficient by the available photovoltaic power. This factor is compared with the first mode threshold. If the load reduction factor is large, the photovoltaic priority mode is selected, resulting in less grid load absorption pressure; otherwise, the energy storage priority mode is selected to absorb surplus power using energy storage. When it is a power absorption identifier, the energy storage rechargeable factor is calculated by multiplying the elasticity coefficient by the energy storage's release power. This factor is compared with the second mode threshold. If the energy storage rechargeable factor is large, the load following mode is selected, absorbing the power gap through flexible loads; otherwise, the grid interaction mode is selected to draw power from the grid. After selecting the main energy dispatch mode, the virtual impedance value and voltage offset coefficient of the dynamic virtual impedance adjustment function are reconfigured. In the photovoltaic priority mode, a low virtual impedance is set, and the voltage offset coefficient is associated with the available photovoltaic power. In the energy storage priority mode, a high virtual impedance is set, and the voltage offset coefficient is associated with the energy storage's release power. In the load following mode, an intermediate virtual impedance is set, and the voltage offset coefficient is associated with the product of the elasticity coefficient and the current voltage. In the grid interaction mode, the virtual impedance is equal to the equivalent impedance of the grid connection interface, and the voltage offset coefficient is zero. The voltage offset is recalculated to replace the initial command, generating a pulse width modulation signal and distributing it to the corresponding converter. Flexible load elasticity is incorporated into the dispatch mode decision logic. When photovoltaic capacity is available, the elasticity coefficient guides photovoltaic or energy storage consumption first; when there is a load gap, flexible load response is prioritized, avoiding unnecessary grid interaction and improving the utilization rate of adjustable resources and energy autonomy on the building side. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the dynamic scheduling method for zero-carbon building photovoltaic-storage-direct-flexible distributed energy systems; Figure 2 This is a flowchart of the multi-protocol communication and data processing of the power adaptive interface of the energy router; Figure 3 This is a flowchart of the initial scheduling instruction generation process for the energy router; Figure 4 This is a flowchart of the arbitration process for switching modes on the energy router. Figure 5 This is a flowchart of the dynamic virtual impedance adjustment function parameter update process; Figure 6 This is a schematic diagram of the pulse width modulation signal distributor structure and signal distribution of an energy router; Figure 7 It is the state of charge distribution curve of a single energy storage battery cell; Figure 8 It is the dynamic change curve of the power demand elasticity coefficient, load reduction factor, and energy storage rechargeable factor. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] See Figure 1 This invention provides a dynamic scheduling method for a zero-carbon building photovoltaic-storage-DC-flexible distributed energy system. The method operates in an energy router within the zero-carbon building, using a power adaptive interface to sense the real-time status of photovoltaic power generation units, energy storage units, and flexible load units in the photovoltaic-storage-DC-flexible network. An initial scheduling command is generated through a DC bus voltage droop control loop embedded with a dynamic virtual impedance adjustment function. The main energy scheduling mode is selected in the mode switching arbitrator by combining the power demand elasticity coefficient. Based on the selected main mode, the parameters of the dynamic virtual impedance adjustment function are corrected and a pulse width modulation signal is generated. Finally, the pulse width modulation signal is distributed to each bidirectional DC converter to complete the dynamic scheduling of the entire distributed energy system.

[0019] Example 1: In specific implementation, please refer to Figure 2 The power adaptive interface of the energy router integrates a first communication protocol stack, a second communication protocol stack, and a third communication protocol stack. The first communication protocol stack establishes a real-time data link with the communication port of the photovoltaic DC-DC converter via a CAN bus. The second communication protocol stack establishes a real-time data link with the communication port of the energy storage battery management system via an RS485 bus. The third communication protocol stack establishes a real-time data link with the communication port of the flexible load controller via an Ethernet interface. The three real-time data links employ an independent cyclic redundancy check mechanism, periodically exchanging data frames at a fixed sampling interval of 200 milliseconds.

[0020] In some implementations, the process of reading the output voltage and current values ​​of the photovoltaic (PV) power generation unit from the communication port of the PV-DC converter is as follows: The power adaptive interface sends a telemetry request frame to the PV-DC converter at a sampling interval of 200 milliseconds. After receiving the telemetry request frame, the PV-DC converter encapsulates the current output voltage and current values ​​of the PV power generation unit into a telemetry response frame and returns it. The power adaptive interface parses the telemetry response frame and extracts the output voltage value from it. and output current value ,in Indicates the sampling time sequence number. (This refers to data from the same sampling time.) and Multiply by the product to obtain the instantaneous photovoltaic power value at that sampling moment. Stored in a circular buffer sorted by time.

[0021] Optionally, the specific method for determining whether a photovoltaic power generation unit is currently in maximum power point tracking state is as follows: extract the instantaneous photovoltaic power value at the adjacent first sampling time from the circular buffer. Instantaneous photovoltaic power value at the second sampling time Calculate the difference .when When the power change direction is recorded, the identifier is "increasing"; when... At that time, the power change direction identifier is recorded as "decreasing". The power change direction identifiers are recorded continuously over three sampling intervals to obtain a direction sequence, such as "increasing-increasing-increasing" or "decreasing-decreasing-decreasing". If the direction sequence consists of three consecutive "increasing" or three consecutive "decreasing" signals, the total power change amplitude over those three sampling intervals is further calculated. The total power change amplitude is... ,in, This represents the instantaneous photovoltaic power value at the end of the third sampling interval. The total power change amplitude is then compared with a preset disturbance amplitude threshold. For comparison, if the change in total power is greater than If the photovoltaic power generation unit is not in maximum power point tracking mode, then it is determined that the total power change is less than or equal to the maximum power point tracking mode. If the direction sequence does not show three consecutive occurrences of the same direction, the photovoltaic power generation unit is determined to be in maximum power point tracking (MPPT) mode. A preset disturbance amplitude threshold is set. Set to 5% of the rated power of the photovoltaic power generation unit. This setting is based on the typical step amplitude of the output power of the photovoltaic module under partial shade.

[0022] In some implementations, the specific process of reading the state of charge (SOC) of each individual energy storage battery cell from the communication port of the energy storage battery management system is as follows: the power adaptive interface sends a cell data request command to the energy storage battery management system through the second communication protocol stack; after receiving the command, the energy storage battery management system sequentially returns the first SOC value of the first energy storage battery cell. The second state of charge value of the second energy storage battery cell Until the The first energy storage battery cell State of charge ,in This represents the total number of individual energy storage battery cells in the energy storage unit. All state-of-charge values ​​are expressed as a percentage with an accuracy of 0.1%.

[0023] The power adaptive interface receives the data from the source. to Mid-scan extraction of maximum value and minimum value Calculate the difference .Will With preset equalization threshold Comparison, preset equalization threshold The setting is 2%, based on the range of single-cell voltage consistency requirements in the national standard for battery management systems. When the energy storage unit's charge balance is at a certain time, it is marked as balanced; when... When the energy storage unit's energy balance status is reached, it is marked as an unbalanced state. The energy balance status mark is stored as a Boolean variable, with a true value indicating a balanced state and a false value indicating an unbalanced state.

[0024] In some implementations, the specific method for reading the allowable power outage duration and adjustable power range of each flexible load from the communication port of the flexible load controller is as follows: the power adaptive interface sends a flexible parameter query message to the flexible load controller through a third communication protocol stack. After parsing the message, the flexible load controller returns the allowable power outage duration and adjustable power range of each flexible load, starting from the first flexible load. The adjustable power range consists of a pair of data consisting of a minimum adjustable power value and a maximum adjustable power value.

[0025] For the first flexible load, the received first permissible power outage duration is denoted as... The first minimum adjustable power value is denoted as The first maximum adjustable power value is denoted as Calculate the first power regulation bandwidth. .Will and Multiplying them together yields the first power demand elasticity coefficient. Repeat the above process to obtain the second power demand elasticity coefficient for the 2nd to Mth flexible loads. To the Power demand elasticity coefficient ,in This represents the total number of flexible loads in the flexible load unit.

[0026] The power demand elasticity coefficient of the flexible load unit is obtained by summing the power demand elasticity coefficients of all flexible loads. The calculation formula is as follows:

[0027] in, The power demand elasticity coefficient of the flexible load unit is expressed in kilowatt-hours. Indicates the total number of flexible loads; Indicates the first The permissible power outage duration for each flexible load, in hours; Indicates the first The power regulation bandwidth of a flexible load, in kilowatts, is calculated as follows: , For the first The maximum adjustable power value of a flexible load. For the first The minimum adjustable power value for a flexible load. The flexible load controller has normalized the power value to kilowatts when returning to the adjustable power range.

[0028] See Figure 7 In the graph, the horizontal axis represents the serial number of the individual energy storage battery cells, ranging from 1 to 500, and the vertical axis represents the state of charge (SOC) of each individual energy storage battery cell, expressed as a percentage (%), ranging from 80% to 100%. The solid curve represents the distribution trend of the SOC of all energy storage battery cells, showing a slow upward trend from low to high, indicating that there are certain differences in the SOC of individual energy storage battery cells, with the lowest SOC being approximately 85.1% and the highest approximately 95.8%. The graph uses dashed lines and dotted lines to indicate the maximum SOC of 95.8% and the minimum SOC of 85.1%, respectively. The dotted line represents the equilibrium threshold δ of 2%, serving as a reference limit for measuring the energy balance of individual energy storage cells.

[0029] The shaded area in the figure corresponds to the range of SOC variation of a single energy storage battery cell, with an amplitude of ΔSOC = 10.6%, which is much higher than the preset equalization threshold δ = 2%. According to the method for determining the state of charge balance of a single energy storage battery cell by the power adaptive interface in Example 1, when the SOC difference ΔSOC of a single energy storage cell is greater than the equalization threshold δ, the energy storage unit's charge balance state is marked as unbalanced.

[0030] The data in this figure reflects the process in this embodiment: after reading the SOC of all individual energy storage cells in the energy storage battery management system through the power adaptive interface of the energy router, the system completes the scanning and difference calculation of the state of charge of the individual cells. This accurately reflects the charge imbalance among the individual energy storage cells, providing a basis for determining the energy balance status of the subsequent energy storage units. Simultaneously, the data analysis results in this figure serve as an important input for determining the energy storage energy balance status, influencing the selection of the main energy dispatch mode and the parameter setting of the dynamic virtual impedance adjustment function, ensuring the safe and stable operation of the energy storage units in the dynamic dispatch method of the distributed energy system.

[0031] Example 2: In specific implementation, please refer to Figure 3 The steps for determining the available photovoltaic power are as follows: First, read the status flag bit of the photovoltaic power generation unit from the status register of the energy router. The value of the status flag bit is written by the power adaptive interface in the process of determining whether the photovoltaic power generation unit is in maximum power point tracking (MPPT) mode. When the value of the status flag bit indicates that the photovoltaic power generation unit is in MPPT mode, the power adaptive interface obtains the output voltage value of the photovoltaic power generation unit at the current sampling time from the communication port of the photovoltaic DC-DC converter. and output current value The product of the two is calculated as the output power value of the photovoltaic power generation unit, and this output power value is assigned to the available photovoltaic power. When the status flag indicates that the photovoltaic (PV) power generation unit is not in maximum power point tracking (MPPT) mode, the power adaptive interface reads the rated power value from the PV power generation unit's nameplate parameter storage area and assigns the rated power value to the available PV power. The nameplate parameter storage area of ​​the photovoltaic power generation unit is located in the non-volatile memory of the energy router and is pre-written by engineers through the human-machine interface.

[0032] The steps for determining the energy storage's discharge capacity are as follows: First, the energy balance status flag bit of the energy storage unit is read from the energy router's status register. The value of this flag bit is written by the power adaptive interface during the process of determining the energy storage unit's energy balance status. When the value of the energy balance status flag bit indicates that the energy storage unit is in a balanced state, the power adaptive interface sends a total discharge capacity query command to the energy storage battery management system via the second communication protocol stack. Upon receiving the total discharge capacity query command, the energy storage battery management system calculates the maximum discharge power that the energy storage unit can continuously output under the current conditions based on the current state of charge, health status, and temperature of all individual energy storage battery cells, and returns this maximum discharge power value as the total discharge capacity parameter. The power adaptive interface then assigns the total discharge capacity parameter to the energy storage's discharge capacity. When the value of the charge balance flag indicates that the energy storage unit is in an unbalanced state, the power adaptive interface sends a minimum charge cell discharge power query command to the energy storage battery management system. After receiving the minimum charge cell discharge power query command, the energy storage battery management system locates the energy storage battery cell with the lowest state of charge from all energy storage battery cells and reads the rated discharge power value of that energy storage battery cell. and current state of charge value The upper limit of the discharge power of a single cell is calculated to be [value missing]. The upper limit of the single-cell discharge power is then returned. The power adaptive interface assigns the upper limit of the single-cell discharge power to the energy storage's discharge power. .

[0033] The energy router integrates an adder module. The two inputs of the adder module are connected to the outputs of the photovoltaic available power register and the energy storage available power register, respectively. The adder module converts the photovoltaic available power... and energy storage power Add them together to obtain the reference power. and reference power Write to the reference power register.

[0034] When generating the initial scheduling command, the dynamic virtual impedance adjustment function reads the preset initial value of the virtual impedance from the parameter memory. and initial value of voltage offset coefficient Virtual impedance initial value This is set as the ratio of one-thousandth of the rated voltage of the DC bus to the rated power of the DC microgrid. In this embodiment, the rated voltage of the DC bus is 750 volts, and the rated power of the DC microgrid is 100 kilowatts. The calculation yields... It is 0.0075 ohms. Initial value of voltage offset coefficient. Setting it to zero ensures that the initial voltage offset is entirely dominated by the virtual impedance. Reference voltage offset. Calculate according to the following formula:

[0035] in, This indicates the reference voltage offset, in volts. This represents the initial value of the virtual impedance, in ohms, and its value is 0.0075. This represents the initial value of the voltage offset coefficient, in volts per kilowatt, and its value is 0. This indicates the reference power, measured in kilowatts.

[0036] Read the base voltage setting value from the voltage setting value memory of the DC bus voltage droop control loop. The voltage setpoint memory of the droop control loop is a dual-port random access memory inside the energy router, containing the base voltage setpoint. This value is written during system initialization; it is the rated DC bus voltage of 750 volts. The energy router uses this base voltage setting. Subtract reference voltage offset The corrected voltage setpoint is obtained. and the corrected voltage setting value Write the voltage setpoint register of the droop control loop. After the voltage setpoint register is updated, a rising edge trigger signal is generated and sent to the droop controller.

[0037] After receiving the rising edge trigger signal, the droop controller reads the corrected voltage setpoint. and the current voltage of the DC bus Calculate voltage deviation The droop controller internally configures droop factors for the photovoltaic bidirectional DC-DC converter, the energy storage bidirectional DC-DC converter, and the load bidirectional DC-DC converter. The droop factor for the photovoltaic bidirectional DC-DC converter... Set as rated power for photovoltaic bidirectional DC-DC converter Deviation from maximum allowable voltage The ratio, Maximum permissible voltage deviation The droop factor is set to 5% of the rated DC bus voltage, i.e., 37.5 volts. This is the droop factor for the energy storage bidirectional DC-DC converter. Set as rated power for energy storage bidirectional DC-DC converter Deviation from maximum allowable voltage The ratio of the load to the droop coefficient of the bidirectional DC-DC converter. Set the rated power of the bidirectional DC-DC converter to the load. Deviation from maximum allowable voltage The ratio. The droop controller calculates the initial duty cycle of the photovoltaic bidirectional DC converter. Initial duty cycle of the energy storage bidirectional DC-DC converter and the initial duty cycle of the bidirectional DC-DC converter ,in , and The rated duty cycles of the photovoltaic bidirectional DC-DC converter, the energy storage bidirectional DC-DC converter, and the load bidirectional DC-DC converter are all set to 0.5.

[0038] Power flow direction indication based on reference power The sign is determined: if If it is a positive number, the power flow direction identifier is assigned the power injection identifier; if If the value is negative, the power flow direction identifier is assigned the power absorption identifier. The energy router sets the initial duty cycle of the photovoltaic bidirectional DC-DC converter. Initial duty cycle of the energy storage bidirectional DC-DC converter Initial duty cycle of the load bidirectional DC-DC converter The power flow direction identifier is packaged into an initial scheduling instruction, which is stored in the instruction buffer as a data structure for the mode switching arbitrator to read.

[0039] Example 3: In specific implementation, please refer to Figure 4 The mode switching arbiter of the energy router contains a preset mode memory, which is a non-volatile read-only memory that stores mode identifiers for four distributed energy collaborative working modes: MODE_PV_PRIORITY for photovoltaic priority mode, MODE_ESS_PRIORITY for energy storage priority mode, MODE_LOAD_FOLLOW for load follow mode, and MODE_GRID_INTERACT for grid interaction mode.

[0040] The mode switching arbitrator reads the initial scheduling instruction from the instruction buffer, parses the data structure of the initial scheduling instruction, and extracts the power flow direction identifier field. The power flow direction identifier field is a Boolean value; a true Boolean value indicates a power injection identifier, and a false Boolean value indicates a power absorption identifier.

[0041] When the power flow direction is identified as a power injection indicator, the mode switching arbitrator enters the power injection processing branch. In this branch, the mode switching arbitrator reads the power demand elasticity coefficient of the flexible load unit from the power demand elasticity coefficient register. Simultaneously, read the available photovoltaic power from the available photovoltaic power register. Calculate the load reduction factor. The calculation method is the power demand elasticity coefficient. The reciprocal of the photovoltaic available power The product of and . Power demand elasticity coefficient The reciprocal of is Available photovoltaic power It is positive when the power injection flag is active. The first mode threshold is located within the mode switching arbitrator. The first mode threshold is pre-configured and stored in the threshold register. The value is 0.8, which is determined based on experience: when the load reduction factor is greater than 0.8, it means that the current available photovoltaic power generation is sufficient to cover most of the power adjustment needs of flexible loads, and photovoltaic power generation should be given priority.

[0042] The mode switching arbitrator will reduce the load by a factor. Compared with the first mode threshold Compare. If If this occurs, a selection signal for the photovoltaic priority mode is generated, setting the current energy dispatching main mode to photovoltaic priority mode; if If this occurs, a selection signal for the energy storage priority mode will be generated, setting the current energy dispatch main mode to the energy storage priority mode.

[0043] When the power flow direction is identified as a power absorption indicator, the mode switching arbitrator enters the power absorption processing branch. In this branch, the mode switching arbitrator reads the power demand elasticity coefficient of the flexible load unit from the power demand elasticity coefficient register. Read the energy storage power from the energy storage power register. Calculate the rechargeability factor of energy storage. The calculation method is the power demand elasticity coefficient. With energy storage and power output The product of the energy storage and its dissipation power. The absolute value is taken under the power absorption flag state and then used in the calculation. The second mode threshold is located within the mode switching arbitrator. The second mode threshold is pre-configured and stored in the threshold register. The value is 50, which is based on the power demand elasticity coefficient. The unit is kilowatt-hour and the energy storage capacity. After multiplying by the unit kilowatt, the energy storage rechargeability factor is... The unit of measurement is kilowatt-hours. When the energy storage rechargeability factor is greater than 50, it indicates that the power elasticity adjustment capability of the flexible load and the charging acceptance capability of the energy storage unit are sufficiently large, and the system should prioritize following load fluctuations.

[0044] The mode switching arbitrator will index the energy storage rechargeability. With the second mode threshold Compare. If If so, a load follow mode selection signal is generated, setting the current energy dispatch master mode to load follow mode; if This generates a grid interaction mode selection signal, setting the current energy dispatch master mode to grid interaction mode.

[0045] In some implementations, after the mode switching arbitrator completes the selection of the primary energy dispatch mode, it writes the selected mode identifier into the primary mode register and generates a mode update interrupt signal to notify the dynamic virtual impedance adjustment function parameter update module to read the contents of the primary mode register. The mode switching arbitrator will never activate two or more primary energy dispatch modes simultaneously in any case, ensuring the consistency of subsequent parameter calculations.

[0046] Energy storage rechargeability index The specific calculation formula is as follows:

[0047] in, The energy storage rechargeability index is a dimensionless value used to quantitatively assess the ability of an energy storage unit to accept excess electricity under current conditions. This represents the power demand elasticity coefficient of the flexible load unit, expressed in kilowatt-hours. This indicates the energy storage capacity, measured in kilowatts (kW). The preset baseline power demand elasticity coefficient, The preset baseline energy storage discharge capacity, and The product is used to normalize the energy storage rechargeability index to a dimensionless value. Second mode threshold. The value of 0.5 is determined based on a large number of simulation experiments. When the energy storage rechargeability index is greater than 0.5, it indicates that the power elasticity adjustment capability of the flexible load and the charging acceptance capability of the energy storage unit are large enough, and the system should prioritize following the load fluctuations.

[0048] As can be seen from the figure, the load reduction factor During the entire 100-second monitoring period, the value fluctuated between approximately 0.65 and 1.3, showing a slow upward trend overall, and consistently remained at the first mode threshold. For most of the time periods mentioned above, it is evident that the inverse product of the available photovoltaic power and the elasticity coefficient of flexible load power demand is relatively large, indicating that the system has a strong photovoltaic power generation utilization capacity.

[0049] Energy storage rechargeable factor It fluctuates in the range of 60 to 85 kilowatt-hours, significantly higher than the threshold of the second mode. The value of kilowatt-hours and its relatively stable fluctuation range reflect that the product of the energy storage unit's dissipable power and the elasticity coefficient of flexible load power demand is always at a high level, indicating that the system has sufficient energy storage charging acceptance capacity.

[0050] Based on the logic of the mode switching arbitrator in Example 3, when the power flow is power injection, the load reduction factor is [not specified]. A value consistently above the threshold of 0.8 indicates that the system prioritizes photovoltaic mode; when the power flow is power absorption, the energy storage rechargeability factor... A value significantly higher than the threshold of 50 indicates that the system prioritizes the load-following mode. The data in the figure demonstrates the real-time response capability of this dynamic scheduling method to the power demand elasticity coefficient and power availability in actual operation, ensuring that the mode switching arbitrator can accurately determine the primary energy dispatch mode that the system should adopt at the moment, and achieving a dynamic balance between prioritizing photovoltaic power generation and energy storage load following.

[0051] Example 4: In specific implementation, please refer to Figure 5 Upon receiving a mode update interrupt signal from the mode switching arbitrator, the dynamic virtual impedance adjustment function parameter update module reads the mode identifier of the current energy dispatching main mode from the main mode register. Internally, the dynamic virtual impedance adjustment function parameter update module contains a parameter configuration table. Indexed by the mode identifier, this table stores the virtual impedance configuration values ​​and voltage offset coefficient configuration logic corresponding to each of the four energy dispatching main modes.

[0052] When the mode identifier in the main mode register is MODE_PV_PRIORITY, indicating that the current energy dispatch main mode is photovoltaic priority mode, the dynamic virtual impedance adjustment function parameter update module sets the virtual impedance value in the function parameters of the dynamic virtual impedance adjustment function to the first impedance value. The first impedance value is 0.05 ohms, which is determined based on the design specifications of the output impedance of the photovoltaic bidirectional DC converter at rated power. The rated power of the photovoltaic bidirectional DC converter is 50 kW, the rated voltage of the DC bus is 750 volts, and the rated current is approximately 66.7 amperes. The first impedance value is taken as the ratio of 0.67 / 1000 of the rated voltage of the DC bus to the rated current, so that the voltage drop is controlled within 5 volts under maximum power point tracking. At the same time, the dynamic virtual impedance adjustment function parameter update module reads the photovoltaic available power from the photovoltaic available power register. Read the rated voltage of the DC bus from the DC bus parameter memory. The voltage offset coefficient in the function parameters of the dynamic virtual impedance adjustment function is set to the photovoltaic available power. With DC bus rated voltage The ratio of the rated voltage of the DC bus. The stored value is 750 volts.

[0053] When the mode identifier in the main mode register is MODE_ESS_PRIORITY, indicating that the current energy dispatch main mode is energy storage priority mode, the dynamic virtual impedance adjustment function parameter update module sets the virtual impedance value in the function parameters of the dynamic virtual impedance adjustment function to the second impedance value. The second impedance value is 0.15 ohms. This value is determined based on three times the margin of the internal resistance characteristic of the energy storage bidirectional DC-DC converter in constant current discharge mode. The rated power of the energy storage bidirectional DC-DC converter is 30 kW, the rated voltage of the DC bus is 750 volts, the rated current is approximately 40 amps, and the internal resistance of the energy storage battery cell is approximately 0.05 ohms. The second impedance value of 0.15 ohms is chosen to provide sufficient damping to suppress voltage oscillations during the energy storage charging and discharging switching process. The second impedance value of 0.15 ohms is greater than the first impedance value of 0.05 ohms. At the same time, the dynamic virtual impedance adjustment function parameter update module reads the energy storage dischargeable power from the energy storage dischargeable power register. Read the rated voltage of the DC bus from the DC bus parameter memory. The voltage offset coefficient in the function parameters of the dynamic virtual impedance adjustment function is set as the energy storage discharge power. With DC bus rated voltage The ratio of .

[0054] When the mode identifier in the main mode register is MODE_LOAD_FOLLOW... the dynamic virtual impedance adjustment function parameter update module reads the power demand elasticity coefficient of the flexible load unit from the power demand elasticity coefficient register. Read the current DC bus voltage from the DC bus voltage sampling module. The voltage offset coefficient in the function parameters of the dynamic virtual impedance adjustment function is set as the power demand elasticity coefficient. Current voltage of DC bus The product is then divided by a preset reference energy value. ,Right now

[0055] in, This is a preset constant, measured in kilowatt-hours, used to ensure the voltage offset coefficient. The dimension is volts per kilowatt.

[0056] When the mode identifier in the main mode register is MODE_GRID_INTERACT, indicating that the current energy dispatch main mode is grid interaction mode, the dynamic virtual impedance adjustment function parameter update module sets the virtual impedance value in the function parameters of the dynamic virtual impedance adjustment function to the fourth impedance value. The value of the fourth impedance value is equal to the equivalent impedance of the grid-connected interface of the energy router. The equivalent impedance of the grid-connected interface of the energy router is measured by an impedance analyzer during the system grid-connected commissioning phase, and the measured value is written into the equivalent impedance register of the grid-connected interface through the human-machine interface. The typical value is 0.08 ohms. This value is determined based on the converted values ​​of the leakage reactance of the grid-connected transformer and the inductive reactance of the AC filter inductor on the DC side. At the same time, the dynamic virtual impedance adjustment function parameter update module sets the voltage offset coefficient in the function parameters of the dynamic virtual impedance adjustment function to zero, so that the DC bus voltage in grid interaction mode is completely determined by the voltage control loop of the grid-connected inverter on the grid side, without any power-related voltage offset.

[0057] After the dynamic virtual impedance adjustment function parameter update module completes the setting of the virtual impedance value and voltage offset coefficient, it records the set virtual impedance value as follows: The set voltage offset coefficient is denoted as Then, the reference voltage offset recalculation process is initiated. The reference voltage offset recalculation process is as follows: the current reference power is read from the reference power register. The reference voltage offset is recalculated using the configured dynamic virtual impedance adjustment function. The calculation formula is as follows:

[0058] in, This represents the recalculated reference voltage offset, in volts. This represents the set virtual impedance value in ohms. Depending on the main energy dispatch mode, the value can be either the first impedance value of 0.05 ohms, the second impedance value of 0.15 ohms, the third impedance value of 0.10 ohms, or the fourth impedance value, which is the equivalent impedance of the grid connection interface. This represents the voltage offset coefficient after setting, which, depending on the main energy dispatch mode, corresponds to the available photovoltaic power. With DC bus rated voltage The ratio of energy storage to power output With DC bus rated voltage The ratio of power demand elasticity coefficient Current voltage of DC bus The product of or zero; This indicates the reference power, measured in kilowatts.

[0059] The dynamic virtual impedance adjustment function parameter update module recalculates the reference voltage offset. Replace the reference voltage offset in the initial scheduling instruction to generate a pulse width modulation signal. The replacement process specifically involves: reading the data structure of the initial scheduling instruction from the instruction buffer, locating the reference voltage offset field, and using... The original values ​​are overwritten, while the power flow direction identification field and the initial duty cycle field remain unchanged. The modified data structure is then written as a pulse width modulation (PWM) signal to the PWM signal generation register. Upon receiving the write operation, the PWM signal generation register generates a ready flag, notifying the PWM signal distributor to begin operation.

[0060] Example 5: In specific implementation, please refer to Figure 6 The pulse width modulation signal distributor of the energy router contains a channel configuration module, which is connected to three independent drive signal generators. These three drive signal generators correspond to the first, second, and third drive channels, respectively. During power-on initialization, the channel configuration module reads the channel binding relationship from the non-volatile configuration memory, establishing an electrical connection between the output of the first drive channel and the gate drive input port of the photovoltaic bidirectional DC-DC converter, the output of the second drive channel and the gate drive input port of the energy storage bidirectional DC-DC converter, and the output of the third drive channel and the gate drive input port of the load bidirectional DC-DC converter. Each of the three drive signal generators includes a duty cycle register and an enable register. The duty cycle register has a 16-bit width, and the enable register is a Boolean register.

[0061] In some implementations, the pulse width modulation (PWM) signal distributor reads the PWM signal data structure from the PWM signal generation register. The PWM signal data structure contains three fields: a power flow direction identifier field, a duty cycle parameter field, and a mode identifier field. The power flow direction identifier field is a Boolean value, the duty cycle parameter field is a 16-bit unsigned integer, and the mode identifier field is a four-bit binary code, corresponding to photovoltaic priority mode, energy storage priority mode, load follower mode, and grid interaction mode, respectively.

[0062] The pulse width modulation signal distributor converts the 16-bit unsigned integer in the duty cycle parameter field into a duty cycle value. The conversion method involves dividing the 16-bit unsigned integer by 65535 to obtain the duty cycle value, which ranges from 0 to 1. .

[0063] The pulse width modulation signal distributor executes channel allocation logic based on the Boolean value of the power flow direction identifier field and the binary code of the mode identifier field.

[0064] When the Boolean value of the power flow direction identifier field is true, meaning the power flow direction identifier is a power injection identifier, the channel configuration module will set the duty cycle value. Simultaneously, the duty cycle registers corresponding to the first driving channel and the second driving channel are written. The write operation is as follows: the channel configuration module uses the internal bus to... The 16-bit binary representation sends the duty cycle register addresses to the first and second drive channels respectively, and a latch signal is sent to both duty cycle registers after writing. Simultaneously, the channel configuration module sets the enable registers of the first and second drive channels to true, enabling their outputs. For the third drive channel, the channel configuration module sets its enable register to false, thus disabling its output. Disabling the third drive channel's output is achieved by setting the output stage of the third drive channel's drive signal generator to a high-impedance state.

[0065] When the Boolean value of the power flow direction identifier field is false, meaning the power flow direction identifier is a power absorption identifier, and the binary code of the mode identifier field corresponds to the load follow mode, the channel configuration module will set the duty cycle value. The duty cycle register corresponding to the third drive channel is written, and the enable register of the third drive channel is set to true. Simultaneously, the channel configuration module sets the enable registers of both the first and second drive channels to false, disabling the output of both drive channels. The output stages of both the first and second drive channels are set to high impedance.

[0066] When the Boolean value of the power flow direction identifier field is false, meaning the power flow direction identifier is a power absorption identifier, and the binary code of the mode identifier field corresponds to the grid interaction mode, the channel configuration module will set the duty cycle value. Simultaneously, the duty cycle registers of the first drive channel, the second drive channel, and the third drive channel are written. After writing is complete, the channel configuration module sets the enable registers of the first drive channel, the second drive channel, and the third drive channel to true, enabling the output of all three drive channels.

[0067] In all three allocation scenarios, after completing the duty cycle value writing and enable register setting, the channel configuration module sends synchronous trigger pulses to the three sets of drive signal generators. Starting from the rising edge of the synchronous trigger pulse, the three sets of drive signal generators generate pulse width modulation (PWM) signals to drive the corresponding bidirectional DC-DC converters based on the values ​​in their respective duty cycle registers and the state of their enable registers. The PWM signal output from the first drive channel is sent to the gate of the power switch of the photovoltaic bidirectional DC-DC converter through an optocoupler isolation drive circuit; the PWM signal output from the second drive channel is sent to the gate of the power switch of the energy storage bidirectional DC-DC converter through an optocoupler isolation drive circuit; and the PWM signal output from the third drive channel is sent to the gate of the power switch of the load bidirectional DC-DC converter through an optocoupler isolation drive circuit.

[0068] See Figure 8 In the graph, the horizontal axis represents time in seconds, and the vertical axis represents duty cycle, ranging from 0 to 0.9. The legend shows that the three curves correspond to the duty cycles of the first drive channel, the second drive channel, and the third drive channel, respectively.

[0069] Within the time interval of 0 to 30 seconds, the duty cycles of the first and second drive channels are stable between approximately 0.45 and 0.53, while the duty cycle of the third drive channel remains at 0. This indicates that the power flow direction indicator during this phase is the power injection indicator. The pulse width modulation signal distributor simultaneously enables the output of the first and second drive channels to drive the photovoltaic bidirectional DC-DC converter and the energy storage bidirectional DC-DC converter, while the drive channel of the load bidirectional DC-DC converter is disabled from output.

[0070] At the 30-second mark, the duty cycles of the first and second drive channels suddenly drop to 0, while the duty cycle of the third drive channel jumps from 0 and fluctuates between approximately 0.48 and 0.56, lasting for approximately 67 seconds. This phenomenon reflects a switch in the power flow indicator from power injection to power absorption, and the mode identifier corresponds to load follow mode. The pulse width modulation signal distributor shuts down the outputs of the first and second drive channels while enabling the output of the third drive channel to drive the load bidirectional DC-DC converter.

[0071] At 67 seconds, the three curves again synchronously jumped to the duty cycle range of approximately 0.50 to 0.60, and the duty cycles of the first, second, and third drive channels became consistent, indicating that the power flow indicator was still power absorption, but the mode identifier switched to grid interaction mode. At this time, the pulse width modulation signal distributor simultaneously enabled the output of all three drive channels, driving the corresponding photovoltaic bidirectional DC-DC converter, energy storage bidirectional DC-DC converter, and load bidirectional DC-DC converter.

[0072] Throughout the test, the duty cycle of each drive channel fluctuated between 0.44 and 0.60, demonstrating that after the dynamic virtual impedance adjustment function parameters were updated, the pulse width modulation signal distributor flexibly completed the enabling and duty cycle allocation of the drive channels according to the logic of different power flow direction identifiers and energy dispatch main mode, ensuring that the power conversion status of each bidirectional DC converter met the system dynamic dispatch requirements.

[0073] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A dynamic scheduling method for a zero-carbon building-based photovoltaic-storage-direct-drive-flexible distributed energy system, characterized in that, The method includes the following steps: In response to voltage fluctuation events on the DC power distribution bus in zero-carbon buildings, the maximum power point tracking status of photovoltaic power generation units, the power balance status of energy storage units, and the power demand elasticity coefficient of flexible load units in the photovoltaic-storage-DC-flexible network are collected through the power adaptive interface of the energy router. Based on the maximum power point tracking status of the photovoltaic power generation unit and the power balance status of the energy storage unit, a dynamic virtual impedance adjustment function is embedded in the DC bus voltage droop control loop to generate an initial scheduling command for balancing power fluctuations in the photovoltaic-storage-DC-flexible network. The initial scheduling command is input to the mode switching arbiter of the energy router, and the power demand elasticity coefficient of the flexible load unit is combined to select the current energy scheduling main mode from a variety of preset distributed energy collaborative working modes. Based on the energy dispatch master mode, the function parameters of the dynamic virtual impedance adjustment function in the DC bus voltage droop control loop are recalculated, and the initial dispatch command is corrected with the function parameters to obtain the pulse width modulation signal for driving the bidirectional DC converter. The pulse width modulation signal is distributed to each bidirectional DC-DC converter corresponding to the photovoltaic power generation unit, the energy storage unit, and the flexible load unit to complete the dynamic scheduling of the zero-carbon building photovoltaic-storage-DC-flexible distributed energy system.

2. The dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 1, characterized in that, Through the power adaptive interface of the energy router, the maximum power point tracking status of photovoltaic power generation units, the power balance status of energy storage units, and the power demand elasticity coefficient of flexible load units in the photovoltaic-storage-DC-flexible network are collected, specifically: In the power adaptive interface of the energy router, real-time data links are established with the communication ports of the photovoltaic DC converter, the energy storage battery management system, and the flexible load controller, respectively. The output voltage and output current of the photovoltaic power generation unit are read from the communication port of the photovoltaic DC converter. By comparing the power change direction at adjacent sampling times, it is determined whether the photovoltaic power generation unit is currently in maximum power point tracking state. The state of charge (SOC) of each individual energy storage battery cell is read from the communication port of the energy storage battery management system. The difference between the maximum and minimum SOC of all energy storage battery cells in the energy storage unit is calculated. Based on the comparison between the difference and a preset equalization threshold, the energy balance state of the energy storage unit is determined. The allowable power outage duration and adjustable power range of each flexible load are read from the communication port of the flexible load controller, and the product of the allowable power outage duration and the adjustable power range is used as the power demand elasticity coefficient of the flexible load unit.

3. The dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 2, characterized in that, Based on the maximum power point tracking status of the photovoltaic power generation unit and the power balance status of the energy storage unit, a dynamic virtual impedance adjustment function is embedded in the DC bus voltage droop control loop to generate an initial scheduling command for balancing power fluctuations in the photovoltaic-storage-DC-flexible network, specifically: When the photovoltaic power generation unit is in maximum power point tracking mode, the output power of the photovoltaic power generation unit is used as the photovoltaic available power; when the photovoltaic power generation unit is not in maximum power point tracking mode, the rated power of the photovoltaic power generation unit is used as the photovoltaic available power. Based on the energy balance state of the energy storage unit, when the difference is less than the preset balance threshold, the total discharge capacity of the energy storage unit is recorded as the energy storage discharge power; when the difference is greater than or equal to the preset balance threshold, the discharge power of the energy storage battery cell with the lowest state of charge in the energy storage unit is taken as the energy storage discharge power. The sum of the available photovoltaic power and the available energy storage power is used as the reference power of the DC bus voltage droop control loop. The reference power is mapped to the reference voltage offset of the DC bus voltage through the dynamic virtual impedance adjustment function. The voltage setting value of the droop control loop is corrected with the reference voltage offset to generate an initial scheduling command for balancing the power fluctuations of the photovoltaic-storage-DC-flexible network.

4. The dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 3, characterized in that, The initial scheduling command is input to the mode switching arbitrator of the energy router. Combined with the power demand elasticity coefficient of the flexible load unit, the primary energy scheduling mode for the current moment is selected from a set of preset distributed energy collaborative working modes. Specifically: In the mode switching arbiter of the energy router, photovoltaic priority mode, energy storage priority mode, load following mode and grid interaction mode are preset as the distributed energy collaborative working modes; The power flow direction identifier in the initial scheduling instruction is parsed as a power injection identifier or a power absorption identifier. When the power flow direction identifier is a power injection identifier, the load reduction factor is calculated based on the product of the reciprocal of the power demand elasticity coefficient and the available photovoltaic power. The load reduction factor is compared with the first mode threshold. If the load reduction factor is greater than the first mode threshold, the photovoltaic priority mode is selected as the main energy scheduling mode at the current moment; otherwise, the energy storage priority mode is selected. When the power flow direction identifier is a power absorption identifier, the energy storage rechargeable factor is calculated based on the product of the power demand elasticity coefficient and the energy storage release power. The energy storage rechargeable factor is compared with the second mode threshold. If the energy storage rechargeable factor is greater than the second mode threshold, the load following mode is selected as the main energy dispatch mode at the current moment; otherwise, the grid interaction mode is selected.

5. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 4, characterized in that, Based on the aforementioned energy dispatch master mode, the function parameters of the dynamic virtual impedance adjustment function in the DC bus voltage droop control loop are recalculated, and the initial dispatch command is corrected using the function parameters to obtain the pulse width modulation signal used to drive the bidirectional DC converter, specifically: When the main energy dispatch mode is photovoltaic priority mode, the virtual impedance value in the function parameters of the dynamic virtual impedance adjustment function is set to the first impedance value, and the voltage offset coefficient in the function parameters of the dynamic virtual impedance adjustment function is set to the ratio of the photovoltaic available power to the DC bus rated voltage. When the main energy dispatch mode is the energy storage priority mode, the virtual impedance value is set to the second impedance value, which is greater than the first impedance value, and the voltage offset coefficient is set to the ratio of the energy storage power to the rated voltage of the DC bus. When the energy dispatching main mode is load following mode, the virtual impedance value is set to the third impedance value, which is between the first impedance value and the second impedance value, and the voltage offset coefficient is set to the product of the power demand elasticity coefficient and the current DC bus voltage. When the energy dispatching main mode is the grid interaction mode, the virtual impedance value is set to the fourth impedance value, which is equal to the equivalent impedance of the grid connection interface of the energy router, and the voltage offset coefficient is set to zero. The reference voltage offset is recalculated using the configured dynamic virtual impedance adjustment function, and the recalculated reference voltage offset is used to replace the reference voltage offset in the initial scheduling instruction to generate the pulse width modulation signal.

6. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 5, characterized in that, The pulse width modulation signal is distributed to each bidirectional DC-DC converter corresponding to the photovoltaic power generation unit, the energy storage unit, and the flexible load unit to complete the dynamic scheduling of the zero-carbon building photovoltaic-storage-DC-flexible distributed energy system. Specifically: In the pulse width modulation signal distributor of the energy router, a first drive channel for the photovoltaic bidirectional DC-DC converter corresponding to the photovoltaic power generation unit, a second drive channel for the energy storage bidirectional DC-DC converter corresponding to the energy storage unit, and a third drive channel for the load bidirectional DC-DC converter corresponding to the flexible load unit are established. According to the power flow direction identifier in the pulse width modulation signal, when the power flow direction identifier is a power injection identifier, the duty cycle parameter in the pulse width modulation signal is simultaneously written into the first driving channel and the second driving channel, and the output of the third driving channel is disabled; According to the power flow direction identifier in the pulse width modulation signal, when the power flow direction identifier is a power absorption identifier and the energy dispatch main mode is a load follow mode, the duty cycle parameter in the pulse width modulation signal is written into the third drive channel, and the output of the first drive channel and the second drive channel is disabled. Based on the power flow direction identifier in the pulse width modulation signal, when the power flow direction identifier is a power absorption identifier and the energy dispatch main mode is the grid interaction mode, the duty cycle parameter in the pulse width modulation signal is simultaneously written into the first drive channel, the second drive channel, and the third drive channel.

7. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 2, characterized in that, The output voltage and output current of the photovoltaic power generation unit are read from the communication port of the photovoltaic DC converter. By comparing the power change direction at adjacent sampling times, it is determined whether the photovoltaic power generation unit is currently in maximum power point tracking state. Specifically: According to a fixed sampling interval, the output voltage and output current values ​​of the photovoltaic power generation unit at multiple sampling times are sequentially read from the communication port of the photovoltaic DC converter. The product of the output voltage value and the output current value at each sampling time is taken as the instantaneous photovoltaic power value at that sampling time. From the photovoltaic instantaneous power values ​​at the multiple sampling times, select the photovoltaic instantaneous power values ​​at the first sampling time and the second sampling time that are adjacent in time, and calculate the difference between the photovoltaic instantaneous power value at the second sampling time and the photovoltaic instantaneous power value at the first sampling time; When the difference is positive, the power change direction is recorded as upward; when the difference is negative, the power change direction is recorded as downward. The power change direction within three consecutive sampling intervals is counted. If the power change direction is upward or downward for three consecutive times and the power change amplitude exceeds the preset disturbance amplitude threshold, it is determined that the photovoltaic power generation unit is not in the maximum power point tracking state; otherwise, it is determined that the photovoltaic power generation unit is in the maximum power point tracking state.

8. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 2, characterized in that, The state of charge (SOC) of each individual energy storage battery cell is read from the communication port of the energy storage battery management system. The difference between the maximum and minimum SOC values ​​of all energy storage battery cells in the energy storage unit is calculated. Based on the comparison between the difference and a preset balancing threshold, the energy balance state of the energy storage unit is determined. Specifically: The first state of charge value of the first energy storage battery cell, the second state of charge value of the second energy storage battery cell, and so on, are read from the communication port of the energy storage battery management system, up to the Nth state of charge value of the Nth energy storage battery cell, where N is the total number of energy storage battery cells in the energy storage unit. Extract the maximum and minimum values ​​from the first state of charge value to the Nth state of charge value, and subtract the minimum value from the maximum value to obtain the difference; The difference is compared with a preset equalization threshold. When the difference is less than the preset equalization threshold, the energy storage unit is marked as being in an equal state. When the difference is greater than or equal to the preset equalization threshold, the energy storage unit is marked as being in an unbalanced state.

9. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 2, characterized in that, The allowable power outage duration and adjustable power range of each flexible load are read from the communication port of the flexible load controller. The product of the allowable power outage duration and the adjustable power range is used as the power demand elasticity coefficient of the flexible load unit. Specifically: The first allowable power outage duration and the first adjustable power range of the first flexible load are read from the communication port of the flexible load controller. The first adjustable power range includes a first minimum adjustable power value and a first maximum adjustable power value. The first power adjustment bandwidth is obtained by subtracting the first minimum adjustable power value from the first maximum adjustable power value. The second allowable power-off duration and the second adjustable power range of the second flexible load are read from the communication port of the flexible load controller, until the Mth allowable power-off duration and the Mth adjustable power range of the Mth flexible load are read, where M is the total number of flexible loads in the flexible load unit, and the power adjustment bandwidth of each flexible load is calculated respectively. The product of the first permissible power outage duration and the first power regulation bandwidth is used as the first power demand elasticity coefficient, and the product of the second permissible power outage duration and the corresponding power regulation bandwidth is used as the second power demand elasticity coefficient, until the Mth power demand elasticity coefficient is obtained. The power demand elasticity coefficients of all flexible loads are summed to obtain the power demand elasticity coefficient of the flexible load unit.

10. A dynamic scheduling method for a zero-carbon building photovoltaic-storage-direct-drive-flexible distributed energy system according to claim 3, characterized in that, The sum of the available photovoltaic power and the available energy storage power is used as the reference power for the DC bus voltage droop control loop. The reference power is mapped to a reference voltage offset of the DC bus voltage using the dynamic virtual impedance adjustment function. This reference voltage offset is then used to correct the voltage setpoint of the droop control loop, generating an initial scheduling command to balance power fluctuations in the photovoltaic-storage-DC-flexible network. Specifically: The available photovoltaic power and the available energy storage power are input into an adder, and the reference power is obtained from the output of the adder. The reference power is input into the dynamic virtual impedance adjustment function, which includes two parameters: virtual impedance value and voltage offset coefficient. The virtual impedance value is multiplied by the reference power to obtain a first intermediate product. The voltage offset coefficient is multiplied by the reference power to obtain a second intermediate product. The first intermediate product and the second intermediate product are added together to obtain the reference voltage offset. The base voltage setting value of the droop control loop is read from the voltage setting value memory of the DC bus voltage droop control loop. The reference voltage offset is subtracted from the base voltage setting value to obtain the corrected voltage setting value. The corrected voltage setting value is written into the voltage setting value register of the droop control loop. The droop control loop is triggered to calculate the initial duty cycle of each bidirectional DC converter based on the corrected voltage setting value. The initial duty cycle is packaged into the initial scheduling command.

Images