Distributed power supply access unit for photovoltaic absorption, intelligent fusion terminal and electric energy meter

Abstract

The invention relates to the technical field of electrical variable measurement, and particularly discloses a photovoltaic absorption-oriented distributed power supply access unit, an intelligent fusion terminal and an electric energy meter. The distributed photovoltaic consumption system is used for solving the problem that the existing distributed photovoltaic consumption system only depends on one-time linkage scheduling and is difficult to consider the multi-time sequence requirements of millisecond-level dynamic support, second-level closed-loop compensation and minute-level economic scheduling, and comprises a distributed power supply access unit, an intelligent fusion terminal and an electric energy meter, according to the invention, on the basis of a closed-loop scheme of multi-level control, double-loop network logic and virtual unit aggregation, rapid response, economic optimization and reliable operation of the high-permeability photovoltaic system are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of electric variable measurement, and in particular to a distributed power supply access unit, an intelligent fusion terminal and an electric energy meter for photovoltaic consumption. Background Art

[0002] The Chinese invention patent with application number 2022112524977 discloses a distribution network and photovoltaic absorption method with distributed power generation penetration. Through the coordinated switching of the dual-loop network structure and the energy storage station, and the linkage scheduling based on the historical power consumption curve and the power generation prediction model, the spatiotemporal matching of photovoltaic output and power demand is achieved, the efficiency of photovoltaic on-site absorption is improved and the loss of abandoned light is reduced. However, the invention only completes the optimization in the one-time linkage scheduling, and has the following shortcomings: In the traditional power operation system, steady-state and transient control are usually divided into primary control (millisecond-level voltage / power support), secondary control (second-minute-level energy balance) and tertiary control (minute-hour-level economic scheduling). The three-level scheduling framework has become an industry practice to ensure the dynamic stability and economy of the power grid. Once the photovoltaic When there is a sudden change in output, the primary control needs to provide instantaneous power support to suppress the voltage drop. The subsequent secondary control needs to dispatch energy storage or adjustable loads in a short time to restore power balance. Finally, the third level of control optimizes the overall grid operation plan and connects it with the market mechanism based on electricity prices, costs and equipment constraints. However, in photovoltaic distributed consumption, in terms of transient support, the use of local energy storage and adjustable loads to provide power or voltage support at the millisecond-second level is not considered, and it still relies on the mechanical action of ring network switching. In terms of short-term energy balance, scheduling is completed only through one-time model prediction + switch switching. There is a lack of closed-loop compensation and rapid correction strategies at the second-minute level. In terms of economic scheduling, it has not been coordinated with intraday / day-ahead market scheduling, and it is impossible to dynamically adjust the output strategy on a minute-hour scale. Therefore, in distributed photovoltaic systems with high penetration and multi-energy coupling, the lack of control strategies that are compatible with the three-level dispatching framework of the power grid not only weakens the ability to quickly respond to sudden power fluctuations, but also makes it difficult to balance long-term economy and real-time stability, seriously restricting the system's dispatching flexibility and operational reliability. It is urgent to propose distributed power access units, intelligent fusion terminals and electricity meters for photovoltaic consumption. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a distributed power access unit, an intelligent fusion terminal and an electric energy meter for photovoltaic consumption, based on a closed-loop solution of multi-level control, dual-loop network logic and virtual unit aggregation, to achieve rapid response, economic optimization and reliable operation of high-penetration photovoltaic systems.

[0004] To achieve the above object, the present invention provides the following technical solutions: The distributed power access unit for photovoltaic consumption includes a fast ring network and an economic ring network topology, and a connecting switch between the fast ring network and the economic ring network. The fast ring network and the economic ring network are both connected to a resource aggregation module. The fast ring network, the economic ring network and the resource aggregation module are all connected to a hierarchical control module. The fast ring network, the economic ring network and the connecting switch are all connected to a fault coordination processing module. The fast ring network is used for millisecond-level primary / secondary control support, and the power flow is reconstructed by switching the two ring network segments to respond to sudden changes in photovoltaic output. The economic ring network is used for three-level economic dispatch at the minute-hour level, and the economic ring network is coordinated with the energy storage station by switching the economic ring network segments to achieve cost-optimized output allocation. The hierarchical control module is used to coordinate the switching sequence of the two ring networks, issue second-level circuit breaking / reclosing instructions to the fast ring network, and issue minute-to-hour-level switch switching and energy storage coordinated scheduling instructions to the economic ring network; The resource aggregation module collects and summarizes the real-time output and status of photovoltaic, energy storage, cooling and heating, and electrical equipment in real time, and aggregates the acquired heterogeneous resources into a unified and controllable virtual unit; The fault collaborative processing module monitors the status of the two ring networks and the contact switch, and when a ring network / equipment abnormality is detected, it achieves millisecond-level isolation through the local switch of the fast ring network, and completes minute-level energy redistribution through the economic ring network and the energy storage station.

[0005] As a further solution of the present invention, the resource aggregation module constructs a Jacobian matrix of the power-voltage equation based on the distribution network power flow model while meeting the set deployment cost budget, and linearizes the Jacobian matrix to obtain the sensitivity coefficient of the power injection change at each bus node to the node voltage and the adjacent branch power flow, and all candidate measuring points are sorted from high to low according to the sensitivity coefficient, and core layer measuring points are arranged at a number of candidate measuring points within the preset ranking number, and the sampling frequency is greater than or equal to 2kHz, and the voltage / current accuracy is ±0.5%FS to respond to the millisecond-level primary / secondary control, and the remaining candidate measuring points are sorted from large to small according to the cost-benefit ratio, and the positions are added in sequence until the cumulative cost is less than or equal to the preset threshold, and these measuring points are identified as auxiliary layer measuring points, and the sampling frequency is greater than or equal to 200Hz and less than 2kHz, and the voltage / current accuracy is ±1%FS. The arrangement method responds to the minute-hour level economic dispatch, and the mobile PMU is deployed on demand in the remaining network area to respond to the third-level control of the minute-hour level economic dispatch with a sampling frequency greater than or equal to 50Hz and less than 200Hz.

[0006] As a further solution of the present invention, the hierarchical control module includes a main control submodule, a secondary control submodule and a tertiary control submodule. The main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it detects that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, it automatically calls the local energy storage device and the adjustable load, calculates and issues millisecond-level transient power support commands through a preset droop control curve, and the droop coefficient used is configured online within 0.02pu-0.05pu to respond to sudden changes in photovoltaic output. The secondary control submodule uses 1s as an update cycle, continuously calculates the average power error in the sliding window of the last 10s, and on this basis generates a power increment instruction through a proportional-integral controller. The proportional gain ranges from 0.1 to 1.0, and the integral gain ranges from 0.01 to 0.1. The command is sent to the energy storage device and the adjustable load through the tie switch to adjust the energy balance in seconds and minutes. The three-level control submodule takes 5 minutes as a scheduling cycle. Based on the cost curve of each distributed resource, an economic optimal scheduling model is established. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-hour scheduling results are converted into closed-loop switch actions and energy storage device charging and discharging instructions for execution.

[0007] As a further solution of the present invention, in the hierarchical control module, the main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it is detected that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, the local energy storage device and the adjustable load are automatically called, and the millisecond-level transient power support command is calculated and issued through the preset droop control curve. The droop coefficient used is configured online within 0.02pu-0.05pu to respond to sudden changes in photovoltaic output; the secondary control submodule uses 1s as the update cycle, continuously calculates the average power error in the sliding window of the last 10s, and on this basis, calculates the average power error through proportional-integral The controller generates a power increment instruction, with a proportional gain value range of 0.1 to 1.0 and an integral gain value range of 0.01 to 0.1. The instruction is sent to the energy storage device and the adjustable load through the interconnecting switch to perform energy balance adjustment at the second-minute level. The three-level control submodule uses a 5-minute scheduling cycle to establish an economically optimal scheduling model based on the cost curve of each distributed resource. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints, and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-hour scheduling results are converted into closed-loop switch actions and energy storage device charging and discharging instructions for execution.

[0008] The intelligent fusion terminal for photovoltaic consumption is connected to the above-mentioned distributed power access unit for photovoltaic consumption, and includes a multi-sensor fusion unit, the multi-sensor fusion unit is connected to an edge preprocessing unit, the edge preprocessing unit is connected to an execution feedback unit, the multi-source sensor fusion unit is used to synchronously collect the operating parameters of the fast ring network and the economic ring network bus side, the edge preprocessing unit is used to perform band-pass filtering, noise suppression, data compression and local caching on the data collected by the multi-source sensor fusion unit at a sampling frequency greater than or equal to 1kHz, the execution feedback unit is used to receive the primary / secondary / tertiary control instructions issued by the distributed power access unit, and control the adjustable load and energy storage device in real time through the local relay / communication interface.

[0009] As a further solution of the present invention, the edge preprocessing unit stores the preprocessed data in a circular buffer greater than or equal to 10MB, encrypts the cached data, and uploads the encrypted cached data in batches after detecting that the communication link with the distributed power supply access unit is restored.

[0010] As a further solution of the present invention, a two-way interaction mechanism is provided between the execution feedback unit and the edge processing unit, including: Parameter delivery: After the execution feedback unit receives the sampling strategy instruction, it writes the new bandpass filter upper and lower limits and sampling frequency meter compression ratio parameters into the edge preprocessing module through the internal bus / communication interface; Status feedback: The edge pre-processing unit dynamically adjusts the signal processing flow based on the received parameters, and reports the buffer usage rate, signal-to-noise ratio of the filtered signal, and data processing delay to the execution feedback unit in real time; Closed-loop verification: The execution feedback unit monitors the buffer utilization rate, signal-to-noise ratio of the filtered signal, and data processing delay of the edge pre-processing unit feedback. If at least one of the following is found, the buffer utilization rate exceeds 90%, the SNR is lower than 20dB, or the processing delay exceeds 10ms, a new sampling frequency and filter bandwidth adjustment instruction will be automatically issued, or a fault recovery plan will be triggered. Timeout protection: The execution feedback unit verifies the response delay of the edge preprocessing unit in a 1s period. If no status feedback is received for three consecutive times, the local backup filtering and compression configuration is started.

[0011] As a further solution of the present invention, the above-mentioned intelligent fusion terminal further includes a self-diagnosis unit, and the self-diagnosis unit is connected to both the execution feedback unit and the edge pre-processing unit.

[0012] As a further solution of the present invention, the workflow of the self-diagnosis unit includes: Step 1, verification signal injection: inject a known verification signal with an amplitude of 1V and a frequency of 1kHz into the edge preprocessing unit at a period of 1Hz; Step 2, output monitoring value: collect the verification signal output after bandpass filtering and compression from the edge preprocessing unit, and measure its amplitude deviation and signal-to-noise ratio; Step 3, abnormality determination and post-processing: When the measured calibration signal amplitude deviation exceeds ±0.5% or the signal-to-noise ratio is lower than 25dB, the preprocessing unit is determined to be abnormal. After the abnormality is determined, the self-diagnosis unit immediately sends a fault alarm to the execution feedback unit, and at the same time instructs the execution feedback unit to switch to the preset backup filter upper and lower limits and compression parameters.

[0013] An electric energy meter comprises at least one of the above-mentioned distributed power access unit for photovoltaic consumption or the above-mentioned intelligent fusion terminal for photovoltaic consumption, measures and reports grid-side electric energy, cooling and heating energy consumption and production load energy data in real time, receives primary / secondary / tertiary hierarchical control instructions and demand response signals issued by the distributed power access unit, and automatically adjusts the start / stop state of local load equipment or controls the charging and discharging action of the energy storage unit according to the received control instructions, so as to realize closed-loop dynamic coupling and optimized consumption of photovoltaic output and multiple loads.

[0014] The technical effects of the distributed power access unit, intelligent fusion terminal and electric energy meter for photovoltaic consumption proposed in the present invention are as follows: The present invention constructs a hierarchical control architecture covering millisecond-level primary / secondary control, second-minute-level secondary energy balance and minute-hour-level three-level economic dispatch, and combines the fast ring network and economic ring network logical topology and interconnecting switches of the dual-ring network to achieve millisecond-level power / voltage support and minute-level economic switching for sudden changes in high-penetration distributed photovoltaic output; and unifies heterogeneous equipment such as photovoltaic, energy storage, cooling and heating, and electricity into controllable virtual units, relying on the two-ring network bus measurement points and adaptive measurement point layout strategy to ensure accurate triggering and rapid closed-loop compensation of primary / secondary control; at the same time, it integrates local cache encryption, execution feedback and self-diagnosis mechanisms to enable the intelligent fusion terminal to achieve real-time response and online correction of dynamic sampling, fault plans and control instructions; and uses numerical algorithms such as the interior point method to update the three-level economic dispatch plan in parallel, complete the coordinated linkage with the day-ahead / intraday market, improve the system's response speed to photovoltaic power fluctuations and voltage stability, optimize the economy of resource dispatch, and enhance the reliability of fault recovery and the flexibility and robustness of the overall operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a block diagram of the distributed power access unit of the present invention; Figure 2 It is a block diagram of the intelligent fusion terminal of the present invention; Figure 3 A physical diagram of the installation of the electric energy meter of the present invention; Figure 4 Statistical table of buffer zone occupancy analysis test data; Figure 5 This is the statistical table of time delay analysis test data; Figure 6 This is the statistical analysis table of the signal-to-noise ratio analysis test data; Figure 7 This is the statistical analysis table of voltage support response test data. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] Example 1 like Figure 1 As shown, the distributed power access unit for photovoltaic consumption proposed in the present invention includes a fast ring network and an economic ring network topology, and a connecting switch located between the two ring networks. The two ring networks are connected to a resource aggregation module, the two ring networks and the resource aggregation module are connected to a hierarchical control module, the two ring networks and the connecting switch are connected to a fault coordination processing module, the fast ring network is used for millisecond-level primary / secondary control support, and the power flow is reconstructed by switching the two ring network segments to respond to sudden changes in photovoltaic output. The economic ring network is used for minute-hour level three-level economic dispatch, and the cost-optimized output distribution is achieved by switching the ring network segments and coordinating with the energy storage station. The hierarchical control module is used to coordinate the switching timing of the two ring networks, issue second-level circuit breaking / reclosing instructions to the fast ring network, and issue minute-to-hour-level switch switching and energy storage coordinated scheduling instructions to the economic ring network. The resource aggregation module collects and summarizes the real-time output and status of photovoltaic, energy storage, heating and cooling, and electrical equipment in real time, and aggregates the acquired heterogeneous resources into a unified controllable virtual unit. The fault collaborative processing module monitors the status of the two ring networks and the interconnecting switches, and when a ring network / equipment abnormality is detected, millisecond-level isolation is achieved through the local switch of the fast ring network, and minute-level energy redistribution is completed through the economic ring network and the energy storage station.

[0018] Compared with the invention patents mentioned in the background technology of the present invention, the present invention has achieved a leap from "one-time linkage" to "second-minute-hour" three-level hierarchical control in the control architecture, shortened the support delay when the photovoltaic output suddenly changes, and enhanced the dynamic response capability to transient power / voltage fluctuations. In view of the radial grid and bidirectional power flow characteristics of the distribution network, the present invention uses the fast ring network logic topology to achieve millisecond-level fault isolation and voltage support, effectively suppressing the busbar overvoltage caused by reverse flow; at the minute-hour level, the economic ring network mode is combined with cost optimization and network power flow safety constraints to complete economic dispatch and line loss minimization. The resource aggregation module uniformly models multi-source heterogeneous equipment such as photovoltaic, energy storage, cold and heat, and adjustable loads as controllable virtual units (visible as virtual power plants in the prior art), which simplifies the dispatch scale and enhances flexibility. The overall solution fully targets the operating difficulties such as voltage over-limit, three-phase imbalance, and excessive harmonic content in the distribution network, and realizes the stable and reliable operation and economical and efficient consumption of high-penetration distributed photovoltaics.

[0019] It should be noted that the resource aggregation module, under the premise of meeting the set deployment cost budget, constructs the Jacobian matrix of the power-voltage equation based on the distribution network power flow model, and linearizes the Jacobian matrix to obtain the sensitivity coefficient of the power injection change at each bus node to the node voltage and the power flow of the adjacent branch. All candidate measuring points are sorted from high to low according to the sensitivity coefficient, and core layer measuring points are arranged at a number of candidate measuring points within the preset ranking number, and the sampling frequency is greater than or equal to 2kHz and the voltage / current accuracy is ±0.5%FS to respond to the millisecond-level primary / secondary control. The remaining candidate measuring points are sorted from large to small according to the cost-benefit ratio, and the positions are added in turn until the cumulative cost is less than or equal to the preset threshold. These measuring points are identified as auxiliary layer measuring points, and the sampling frequency is greater than or equal to 200Hz and less than 2kHz, and the voltage / current accuracy is ±1%FS. The arrangement method responds to the minute-hour economic dispatch, and the mobile PMU is deployed on demand in the remaining network areas to respond to the third-level control of the minute-hour economic dispatch with a sampling frequency greater than or equal to 50Hz and less than 200Hz.

[0020] The use of optimized measurement point layout and adaptive monitoring based on power flow sensitivity greatly improves the observability of the network, and can accurately locate faults and quickly compensate in a closed loop. Under the premise of meeting the deployment cost budget, based on the linearized sensitivity coefficient of the Jacobian matrix of the distribution network power flow, the core layer measurement points and auxiliary layer measurement points are hierarchically constructed and supplemented by mobile PMUs, so as to achieve high-precision and low-latency observation of the primary / secondary control triggering of the fast ring network at the millisecond-second level and the three-level economic dispatch of the economic ring network at the minute-hour level. At the same time, the network observability and fault location accuracy are greatly improved, which not only ensures the rapid response to the sudden change of photovoltaic output and the stability of bus voltage, but also supports the coordination of hierarchical economic dispatch and precise energy storage. While reducing the deployment cost of measurement points and the pressure of communication bandwidth, it realizes the multi-sequence and multi-energy coordinated closed-loop efficient operation for high-penetration distributed photovoltaic systems.

[0021] For example, in a distribution network fast ring network, the total length is about 4km, with a total of 16 feeder branch nodes, and each node is a candidate measurement point. The total cost corresponding to the measurement point deployment budget does not exceed 100,000 yuan. The resource aggregation module first constructs the power-voltage Jacobian matrix based on the power flow model of the distribution network, and calculates the sensitivity coefficient of the injected power change of each node to the node voltage and the adjacent branch power flow. Among them, nodes 3, 7, 10, and 14 have the highest sensitivity; select the top 4 nodes 3, 7, 10, and 14 with the highest sensitivity, and install voltage / current sensors with an accuracy of ±0.5%FS respectively. The sampling rate is set to 2kHz. In the scenario of a sudden drop in photovoltaic output, it only takes 5ms to detect a voltage drop of 2% at node 7. , immediately triggering the local energy storage device to inject 50kW power into the bus, quickly restoring the voltage to above 0.98pu; among the remaining 12 candidate nodes, they were sorted from high to low according to the sensitivity coefficient and cost ratio, and 5 nodes, 2, 5, 8, 12, and 15, were selected in succession. The sampling rate was set to 500Hz and ±1%FS. During minute-level economic dispatch, the predicted load error was reduced from the original ±8% to ±2% through the measurement point data of nodes 5 and 12, significantly improving the accuracy and economy of energy storage dispatch; 2 mobile PMUs were deployed at the remaining nodes, each with a coverage radius of 500m and a sampling rate of 100Hz, for short-term fault drills and blind spot observations. When a fault occurred between lines 12-13, the mobile PMU completed the fault location within 3s, and assisted the rapid ring network to isolate the fault branch section, reducing voltage disturbances in non-fault areas.

[0022] It should be noted that the hierarchical control module includes a main control submodule, a secondary control submodule and a tertiary control submodule. The main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it detects that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, it automatically calls the local energy storage device and the adjustable load, and calculates and issues millisecond-level transient power support commands through the preset droop control curve. The droop coefficient used is configured online within 0.02pu-0.05pu to respond to sudden changes in photovoltaic output. The secondary control submodule uses 1s as an update cycle, continuously calculates the average power error in the sliding window of the last 10s, and on this basis generates a power increment command through a proportional-integral controller. The proportional gain ranges from 0.1 to 1.0, and the integral gain ranges from 0.01 to 0.1. The command is sent to the energy storage device and the adjustable load through the tie switch to adjust the energy balance in seconds and minutes. The three-level control submodule takes 5 minutes as a scheduling cycle. Based on the cost curve of each distributed resource, an economic optimal scheduling model is established. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-hour scheduling results are converted into closed-loop switch actions and energy storage device charging and discharging instructions for execution.

[0023] In the scenario of high-penetration distributed photovoltaic grid-connected power, since photovoltaic output is severely affected by meteorological conditions, its power fluctuations include both millisecond-level transient disturbances, as well as second-minute-level short-term deviations and minute-hour-level load / electricity price dynamic changes. Traditional single-level dispatch (one-time model prediction + ring network switching) is difficult to take into account the response requirements of different time scales, resulting in support delays in the system when responding to sudden large output fluctuations, and it is difficult to quickly adjust strategies and synchronize with market signals at the economic dispatch level. Therefore, the hierarchical control architecture covering the main (primary / secondary) control submodule, the secondary control submodule and the tertiary control submodule is set up to integrate the millisecond-level, second-minute-level and minute-hour-level control rhythms in accordance with the multi-time series characteristics of the dynamic coupling of distributed photovoltaic output and distribution network load. The main control submodule uses ≥2kHz high-speed sampling and online adjustable droop coefficient to immediately issue transient power support instructions when the voltage deviation is ±1% or the frequency deviation is ±0.02Hz, reducing the support delay to <10ms, effectively stabilizing the bus voltage and preventing voltage drops caused by reverse flow; the secondary control submodule uses 1s update, 10s sliding window and PI control to accurately allocate energy storage and adjustable loads, complete the balance within 5-15s after the short-term power deviation occurs, and greatly reduce the frequency fluctuation caused by energy mismatch; the tertiary control submodule uses a 5min scheduling cycle, combined with each resource cost curve and network flow safety constraints, to quickly generate the optimal scheduling plan through the interior point method, and automatically switch the loop switch and energy storage charging and discharging to ensure the minimum operating cost and the lowest line loss under dynamic electricity prices and load changes. This hierarchical control strategy targets the multi-sequence dynamic characteristics of the distribution network, which not only improves the system's rapid response capability, but also takes into account the economy and safety of long-term operation.

[0024] It should be specifically explained that in the hierarchical control module, the main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it is detected that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, it automatically calls the local energy storage device and the adjustable load, and calculates and issues millisecond-level transient power support commands through the preset droop control curve. The droop coefficient used is configured online within 0.02pu-0.05pu to respond to sudden changes in photovoltaic output. The secondary control submodule uses 1s as the update cycle, continuously calculates the average power error in the sliding window of the last 10s, and on this basis, uses proportional-integral control. The generator generates a power increment instruction, with a proportional gain value range of 0.1 to 1.0 and an integral gain value range of 0.01 to 0.1. The instruction is sent to the energy storage device and the adjustable load through the tie switch to perform second-minute energy balance adjustment; the three-level control submodule uses 5 minutes as a scheduling cycle, and establishes an economic optimal scheduling model based on the cost curve of each distributed resource. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-hour scheduling results are converted into closed-loop switch actions and energy storage device charging and discharging instructions for execution.

[0025] This hierarchical control strategy significantly improves the dynamic response speed and economic dispatch efficiency of high-penetration distributed photovoltaic systems: First, the main control submodule achieves millisecond-level precise support for the voltage / frequency on the bus side of the fast ring network through ultra-high frequency real-time monitoring of ≥2kHz and an online configurable 0.02pu-0.05pu droop coefficient. It can suppress the voltage drop to within ±0.5% and restore the frequency deviation to within ±0.01Hz within 10ms after a sudden change in photovoltaic output, thereby avoiding protection malfunction and voltage instability caused by reverse flow or transient disturbances; Secondly, the secondary control submodule uses the PI algorithm with a 1s update cycle and a 10s sliding window, which can narrow the power deviation from ±10% to ±2% within 20s, realize fast closed-loop compensation for energy storage and adjustable loads, and significantly reduce system frequency oscillation and energy mismatch; finally, the tertiary control submodule uses a 5min dispatch cycle based on the interior point method optimization of each resource cost curve and network flow safety constraints, which not only seamlessly connects with the day-ahead / intraday market signals, but also reduces the overall operating cost by about 5% and reduces line losses by about 8%, taking into account the balance of supply and demand and economy. This solution effectively solves the contradiction between traditional one-time dispatching in fast response, fault recovery and economic optimization through primary-secondary-tertiary multi-sequence coordination, and comprehensively enhances the stability, reliability and economy of the distribution network under high penetration photovoltaic conditions.

[0026] Example 2 The difference between Example 2 of the present invention and Example 1 is that this example introduces an intelligent fusion terminal for photovoltaic consumption.

[0027] like Figure 2 As shown, the intelligent fusion terminal for photovoltaic consumption proposed in the present invention is connected to the above-mentioned distributed power access unit for photovoltaic consumption, and includes a multi-sensor fusion unit, the multi-sensor fusion unit is connected to an edge preprocessing unit, the edge preprocessing unit is connected to an execution feedback unit, the multi-source sensor fusion unit is used to synchronously collect the operating parameters of the fast ring network and the economic ring network bus side, the edge preprocessing unit is used to perform band-pass filtering, noise suppression, data compression and local caching on the data collected by the multi-source sensor fusion unit at a sampling frequency greater than or equal to 1kHz, the execution feedback unit is used to receive the primary / secondary / tertiary control instructions issued by the distributed power access unit, and control the adjustable load and energy storage device in real time through the local relay / communication interface.

[0028] By integrating multi-source sensing, edge preprocessing and local execution feedback, the intelligent fusion terminal of the present invention can achieve "full-link" low-latency and autonomous closed-loop control on site. Specific technical effects include: High-fidelity and fast response: The multi-sensor fusion unit synchronously captures the voltage, current, temperature and other operating parameters of the fast ring network and the economic ring network at an ultra-high sampling rate of ≥1kHz, ensuring real-time perception of "zero blind zone" of millisecond-level power fluctuations; Local intelligent filtering and data compression: The edge preprocessing unit performs bandpass filtering, noise suppression and data compression locally, effectively eliminating useless information and reducing communication volume by more than 50%, which not only improves measurement accuracy but also reduces the network load of the superior dispatch center; Autonomous closed-loop execution: After receiving the primary / secondary / tertiary control instructions, the execution feedback unit sends them to the adjustable load and energy storage device through the local relay or communication interface in milliseconds, without the need to frequently rely on remote central control, significantly shortening the control delay and enhancing the local autonomy when communication is interrupted; Enhanced disturbance resistance and reliability: The local cache mechanism ensures data integrity during disconnection, and the execution feedback and self-diagnosis work together to automatically switch to the backup strategy when sensing or communication anomalies are detected, ensuring continuous and stable operation of the system under high penetration photovoltaic fluctuations and network interference.

[0029] It should be noted that the edge preprocessing unit stores the preprocessed data in a circular buffer greater than or equal to 10MB, and encrypts the cached data. After detecting that the communication link with the distributed power access unit is restored, the encrypted cached data is uploaded in batches.

[0030] There is often network jitter or link interruption at the power distribution site. A 10MB or larger circular buffer can be used to cache high-frequency pre-processed data locally for hours or even days, ensuring that critical operating information is not lost after the network is restored. The cached data is encrypted at the AES-256 level to prevent tampering or theft during local storage or batch upload, meeting the strict requirements of operation and maintenance and compliance for data privacy and confidentiality. After the link is restored, data is uploaded in batches, which can fully utilize the available bandwidth and avoid network congestion and dispatch center processing pressure caused by frequent small packet transmission. The circular buffer combined with batch upload can retain the complete order and timestamp of the data, which is conducive to subsequent abnormal diagnosis, fault backtracking and accurate energy balance analysis.

[0031] It should be specifically noted that a two-way interaction mechanism is provided between the execution feedback unit and the edge processing unit, including: Parameter delivery: After the execution feedback unit receives the sampling strategy instruction, it writes the new bandpass filter upper and lower limits and sampling frequency meter compression ratio parameters into the edge preprocessing module through the internal bus / communication interface; Status feedback: The edge pre-processing unit dynamically adjusts the signal processing flow based on the received parameters, and reports the buffer usage rate, signal-to-noise ratio of the filtered signal, and data processing delay to the execution feedback unit in real time; Closed-loop verification: The execution feedback unit monitors the buffer utilization rate, signal-to-noise ratio of the filtered signal, and data processing delay of the edge pre-processing unit feedback. If at least one of the following is found, the buffer utilization rate exceeds 90%, the SNR is lower than 20dB, or the processing delay exceeds 10ms, a new sampling frequency and filter bandwidth adjustment instruction will be automatically issued, or a fault recovery plan will be triggered. Timeout protection: The execution feedback unit verifies the response delay of the edge preprocessing unit in a 1s period. If no status feedback is received for three consecutive times, the local backup filtering and compression configuration is started.

[0032] By sending parameters, the edge preprocessing module can switch the bandpass filter bandwidth, sampling frequency and compression ratio in real time according to different output fluctuations and network environments, so as to maintain the optimal filtering and data accuracy when high-frequency jitter or low-frequency trend changes. State feedback ensures that the execution feedback unit is always aware of key indicators such as buffer utilization, signal-to-noise ratio (SNR) and processing delay. When any indicator exceeds the limit, the system can automatically adjust the sampling and filtering parameters at the millisecond-second level, or trigger a fault recovery plan to ensure data flow continuity and measurement reliability. The timeout protection mechanism can quickly switch to the preset backup filtering and compression configuration when the link or module is abnormal, avoiding data leakage or control command delay caused by communication interruption or hardware jamming, and enhancing on-site autonomy.

[0033] In a laboratory environment, simulate the periodic packet loss and delay jitter of the IEC61850 / MQTT link, record the buffer occupancy, SNR and delay change curves of the edge preprocessing module under different sampling strategy instructions; test the recovery time and amplitude of the performance indicators after the closed-loop verification is triggered by parameter adjustment. Apply step-type load increase and decrease disturbances to the fast ring network, send different droop coefficients, sampling frequencies and filter bandwidths in real time, evaluate the system's support delay and recovery accuracy for voltage / frequency deviations, and verify the role of the adaptive sampling strategy in primary / secondary control. Intentionally shut down the edge preprocessing unit communication or simulate the processor overload scenario, observe the behavior of the execution feedback unit switching to the backup parameters under the timeout protection condition, and evaluate the data continuity and measurement accuracy after switching. Figure 4-Figure 7 The data shown shows that under the conditions of simulating 15% packet loss and an average round-trip delay jitter of 100ms, the cache peak occupancy is 92%. After closed-loop verification, the sampling frequency and filter bandwidth are adjusted to reduce the peak to 75%, and the average is maintained below 65%. In the initial state, the SNR of the filtered signal is 15dB; after the new bandpass filter parameters are issued, the SNR is increased to 28dB (an increase of 13dB). The average end-to-end delay of the edge preprocessing module is reduced from 22ms to 7ms, and the peak delay is reduced from 45ms to 12ms. When a ±200 kW step power disturbance is applied to the fast ring network bus, a support command is issued within 5ms after the voltage drops by 0.05pu (i.e. 5%), and the voltage is restored to 0.99pu (origin 1.00pu) within 10ms. The disturbance causes a frequency deviation of ±0.04Hz, and the main control submodule recovers to the range of ±0.01Hz within 8ms; the secondary control reduces the power deviation from the initial ±12% to ±2% within 12s. The ring buffer can cache ≥30,000 samples (corresponding to 10MB), and no samples are missed during the 3s network disconnection. After the execution feedback unit has no state feedback for 3 consecutive seconds, it completes the issuance and loading of backup filtering / compression parameters for the edge preprocessing module within 2 seconds. After the connection is restored, the encrypted data uploaded in batches is decrypted and compared, and the timing error between the beginning and the end is less than 1ms, and the packet loss rate is 0.

[0034] It should be noted that the above-mentioned intelligent fusion terminal also includes a self-diagnosis unit, which is connected to both the execution feedback unit and the edge pre-processing unit. The working process of the self-diagnosis unit includes: Step 1, verification signal injection: inject a known verification signal with an amplitude of 1V and a frequency of 1kHz into the edge preprocessing unit at a period of 1Hz; Step 2, output monitoring value: collect the verification signal output after bandpass filtering and compression from the edge preprocessing unit, and measure its amplitude deviation and signal-to-noise ratio; Step 3, abnormality determination and post-processing: When the measured calibration signal amplitude deviation exceeds ±0.5% or the signal-to-noise ratio is lower than 25dB, the preprocessing unit is determined to be abnormal. After the abnormality is determined, the self-diagnosis unit immediately sends a fault alarm to the execution feedback unit, and at the same time instructs the execution feedback unit to switch to the preset backup filter upper and lower limits and compression parameters.

[0035] By regularly injecting known calibration signals and monitoring their filtered output, the performance degradation or module failure of the edge preprocessing unit can be accurately identified within a 1Hz rhythm, avoiding long-term blind zone operation. Once the amplitude deviation exceeds ±0.5% or the SNR is lower than 25dB, the fault alarm is immediately triggered and the execution feedback unit is instructed to switch to the preset backup filtering and compression parameters, so that the continuity of the data processing link can be maintained without manual intervention. Through real-time calibration and automatic switching, it is ensured that the measurement data uploaded to the distributed power access unit always meets the accuracy and signal-to-noise ratio requirements, providing reliable and high-fidelity basic data for subsequent primary / secondary / tertiary control. The self-diagnosis and automatic switching mechanism greatly shortens the fault location and recovery time, reduces the frequency of on-site maintenance, effectively reduces the operation and maintenance costs, and enhances the anti-disturbance capability in complex power grid environments. Combined with multi-level closed-loop control, the real-time health monitoring and automatic recovery capabilities of the self-diagnosis unit further improve the overall stability and robustness of the distribution network when the photovoltaic output fluctuates violently.

[0036] Example 3 Different from Embodiment 1 and Embodiment 2, this embodiment specifically introduces an electric energy meter proposed by the present invention.

[0037] like Figure 3 As shown, an electric energy meter proposed in the present invention is installed in an electric energy meter box, including at least one of the above-mentioned distributed power access unit for photovoltaic consumption or the above-mentioned intelligent fusion terminal for photovoltaic consumption, real-time measurement and reporting of grid-side electric energy, cold and hot energy consumption and production load energy data, receiving primary / secondary / tertiary level control instructions and demand response signals issued by the distributed power access unit, and automatically adjusting the start and stop status of local load equipment or controlling the charging and discharging actions of the energy storage unit according to the received control instructions, so as to realize the closed-loop dynamic coupling and optimized consumption of photovoltaic output and multiple loads.

[0038] In summary, according to the technical solutions of the present invention described in Examples 1 to 3, the present invention constructs a hierarchical control architecture covering millisecond-level primary / secondary control, second-minute-level secondary energy balance and minute-hour-level three-level economic dispatch, and combines the fast ring network and economic ring network logical topology and tie switches of the dual-ring network to achieve millisecond-level power / voltage support and minute-level economic switching for sudden changes in high-penetration distributed photovoltaic output; and unifies heterogeneous equipment such as photovoltaic, energy storage, cold and heat, and electricity into controllable virtual units, relying on the two-ring network bus measurement points and adaptive measurement point layout strategy to ensure accurate triggering and rapid closed-loop compensation of primary / secondary control; at the same time, integrates local cache encryption, execution feedback and self-diagnosis mechanisms, so that the intelligent fusion terminal can achieve real-time response and online correction of dynamic sampling, fault plans and control instructions; and uses numerical algorithms such as the interior point method to update the three-level economic dispatch scheme in parallel, completes the coordinated linkage with the day-ahead / intraday market, improves the system's response speed to photovoltaic power fluctuations and voltage stability, optimizes the economy of resource scheduling, and enhances the reliability of fault recovery and the flexibility and robustness of the overall operation of the distribution network.

[0039] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0040] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A distributed power access unit for photovoltaic consumption, characterized in that: It includes a fast ring network and an economic ring network topology, and a connecting switch located between the fast ring network and the economic ring network. The fast ring network and the economic ring network are both connected to a resource aggregation module. The fast ring network, the economic ring network and the resource aggregation module are all connected to a hierarchical control module. The fast ring network, the economic ring network and the connecting switch are all connected to a fault coordination processing module. The fast ring network is used for millisecond-level primary / secondary control support, and the power flow is reconstructed by switching the two ring network segments to respond to sudden changes in photovoltaic output. The economic ring network is used for minute-hour level three-level economic dispatch, and the economic ring network is coordinated with the energy storage station by switching the economic ring network segments to achieve cost-optimized output allocation. The hierarchical control module is used to coordinate the switching sequence of the two ring networks, issue second-level circuit breaking / reclosing instructions to the fast ring network, and issue minute-to-hour-level switch switching and energy storage coordinated scheduling instructions to the economic ring network; The resource aggregation module collects and summarizes the real-time output and status of photovoltaic, energy storage, cooling and heating, and electrical equipment in real time, and aggregates the acquired heterogeneous resources into a unified and controllable virtual unit; The fault collaborative processing module monitors the status of the two ring networks and the tie switch, and when a ring network / equipment abnormality is detected, it achieves millisecond-level isolation through the local switch of the fast ring network, and completes minute-level energy redistribution through the economic ring network and the energy storage station.

2. The distributed power access unit for photovoltaic consumption according to claim 1, characterized in that: The resource aggregation module constructs the Jacobian matrix of the power-voltage equation based on the distribution network power flow model while meeting the set deployment cost budget, and linearizes the Jacobian matrix to obtain the sensitivity coefficient of the power injection change at each bus node to the node voltage and the power flow of the adjacent branch. All candidate measuring points are sorted from high to low according to the sensitivity coefficient, and core layer measuring points are arranged at a number of candidate measuring points within the preset ranking number, and the sampling frequency is greater than or equal to 2kHz and the voltage / current accuracy is ±0.5%FS to respond to the millisecond-level primary / secondary control. The remaining candidate measuring points are sorted from large to small according to the cost-benefit ratio, and the positions are added in sequence until the cumulative cost is less than or equal to the preset threshold. These measuring points are identified as auxiliary layer measuring points, and the sampling frequency is greater than or equal to 200Hz and less than 2kHz, and the voltage / current accuracy is ±1%FS to respond to the minute-hour economic dispatch. In the remaining network areas, mobile PMUs are deployed as needed to respond to the third-level control of the minute-hour economic dispatch with a sampling frequency greater than or equal to 50Hz and less than 200Hz.

3. The distributed power access unit for photovoltaic consumption according to claim 1, characterized in that: The hierarchical control module includes a main control submodule, a secondary control submodule and a tertiary control submodule; the main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it detects that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, it automatically calls the local energy storage device and the adjustable load, and calculates and issues millisecond-level transient power support commands through the preset droop control curve. The droop coefficient used is configured online within 0.02pu-0.05pu to respond to sudden changes in photovoltaic output. The secondary control submodule uses 1s as the update cycle, continuously calculates the average power error in the sliding window of the last 10s, and on this basis, The proportional-integral controller generates a power increment command, with the proportional gain ranging from 0.1 to 1.0 and the integral gain ranging from 0.01 to 0.

1. The command is sent to the energy storage device and the adjustable load through the tie switch to perform energy balance adjustment at the second-minute level. The three-level control submodule uses 5 minutes as a scheduling cycle and establishes an economically optimal scheduling model based on the cost curve of each distributed resource. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-hour scheduling results are converted into closed-loop switch actions and energy storage device charging and discharging instructions for execution.

4. The distributed power access unit for photovoltaic consumption according to claim 3 is characterized in that: In the hierarchical control module, the main control submodule monitors the voltage and frequency on the bus side of the fast ring network in real time at a sampling frequency greater than or equal to 2kHz. When it is detected that the frequency deviation exceeds ±0.02Hz / the voltage deviation exceeds ±1%, the local energy storage device and the adjustable load are automatically called, and the millisecond-level transient power support command is calculated and issued through the preset droop control curve. The droop coefficient used is configured online within 0.02pu-0.05pu to respond to the sudden change of photovoltaic output; the secondary control submodule takes 1s as the update cycle, continuously calculates the average power error in the sliding window of the last 10s, and generates a power increment command through the proportional-integral controller on this basis. The proportional gain ranges from 0.1 to 1.0, and the integral gain ranges from 0.01 to 0.

1. The command is issued to the energy storage device and the adjustable load through the tie switch to perform energy balance adjustment at the second-minute level; The three-level control submodule uses a 5-minute scheduling cycle and establishes an economically optimal scheduling model based on the cost curve of each distributed resource. Under the premise of ensuring that the total output is equal to the predicted load, the output of each resource meets the upper and lower limit constraints and the network flow is safe, the cost minimization problem is solved through numerical algorithms such as the interior point method, and the minute-to-hour scheduling results are converted into closing-loop switch actions and energy storage device charging and discharging instructions for execution.

5. Intelligent fusion terminal for photovoltaic consumption, characterized by: A distributed power access unit for photovoltaic consumption connected to any one of claims 1 to 4, and including a multi-sensor fusion unit, the multi-sensor fusion unit is connected to an edge preprocessing unit, the edge preprocessing unit is connected to an execution feedback unit, the multi-source sensor fusion unit is used to synchronously collect the operating parameters of the fast ring network and the economic ring network bus side, the edge preprocessing unit is used to perform band-pass filtering, noise suppression, data compression and local caching on the data collected by the multi-source sensor fusion unit at a sampling frequency greater than or equal to 1kHz, the execution feedback unit is used to receive the primary / secondary / tertiary control instructions issued by the distributed power access unit, and control the adjustable load and energy storage device in real time through the local relay / communication interface.

6. The intelligent fusion terminal for photovoltaic consumption according to claim 5 is characterized in that: The edge preprocessing unit stores the preprocessed data in a circular buffer greater than or equal to 10MB and encrypts the cached data. After detecting that the communication link with the distributed power access unit is restored, the encrypted cached data is uploaded in batches.

7. The intelligent fusion terminal for photovoltaic consumption according to claim 6 is characterized in that: There is a two-way interaction mechanism between the execution feedback unit and the edge processing unit, including: Parameter delivery: After the execution feedback unit receives the sampling strategy instruction, it writes the new bandpass filter upper and lower limits and sampling frequency meter compression ratio parameters into the edge preprocessing module through the internal bus / communication interface; Status feedback: The edge pre-processing unit dynamically adjusts the signal processing flow based on the received parameters, and reports the buffer usage rate, signal-to-noise ratio of the filtered signal, and data processing delay to the execution feedback unit in real time; Closed-loop verification: The execution feedback unit monitors the buffer utilization rate, signal-to-noise ratio of the filtered signal, and data processing delay of the edge pre-processing unit feedback. If at least one of the following is found, the buffer utilization rate exceeds 90%, the SNR is lower than 20dB, or the processing delay exceeds 10ms, a new sampling frequency and filter bandwidth adjustment instruction will be automatically issued, or a fault recovery plan will be triggered. Timeout protection: The execution feedback unit verifies the response delay of the edge preprocessing unit in a 1s period. If no status feedback is received for three consecutive times, the local backup filtering and compression configuration is started.

8. The intelligent fusion terminal for photovoltaic consumption according to claim 5 is characterized in that: It also includes a self-diagnosis unit, which is connected to the execution feedback unit and the edge pre-processing unit.

9. The intelligent fusion terminal for photovoltaic consumption according to claim 8 is characterized in that: The workflow of the self-diagnosis unit includes: Step 1, verification signal injection: inject a known verification signal with an amplitude of 1V and a frequency of 1kHz into the edge preprocessing unit at a period of 1Hz; Step 2, output monitoring value: collect the verification signal output after bandpass filtering and compression from the edge preprocessing unit, and measure its amplitude deviation and signal-to-noise ratio; Step 3, abnormality determination and post-processing: When the measured calibration signal amplitude deviation exceeds ±0.5% or the signal-to-noise ratio is lower than 25dB, the preprocessing unit is determined to be abnormal. After the abnormality is determined, the self-diagnosis unit immediately sends a fault alarm to the execution feedback unit, and at the same time instructs the execution feedback unit to switch to the preset backup filter upper and lower limits and compression parameters.

10. An electric energy meter, characterized in that: The method comprises a distributed power access unit for photovoltaic absorption as described in any one of claims 1 to 4, which measures and reports grid-side electric energy, cooling and heating energy consumption and production load energy data in real time, receives primary / secondary / tertiary level control instructions and demand response signals issued by the distributed power access unit, and automatically adjusts the start / stop state of local load equipment or controls the charging and discharging action of the energy storage unit according to the received control instructions, so as to realize closed-loop dynamic coupling and optimized absorption of photovoltaic output and multiple loads.

Images